1870 E N Welch alarm clock

[Simon Holt]

I posted a query about this clock originally in General Clock Discussions (E N Welch Steeple Alarm clock: looking for date), in order to get it dated. I'm posting now in Clock Repair for guidance on a repair issue. If it's better to have one thread for one clock (regardless of the nature of the query) then maybe an admin would like to merge the two threads.

This clock came to me with a request to 'get it running again'. I don't know how many 150-year-old clocks there are out there that still belong to the original owner, but this is one!

Initial triage showed that, apart from being gummed with old oil and missing its alarm stop plate, the teeth on wheel of T1 show the following combination of wear and distortion:

The corresponding lantern pinion does not to seem too bad:

Before cleaning, the clock runs if the spring is more than half-wound. But does that wheel look close to causing terminal failure? Is it a suitable candidate for planishing?

I've seen that some people deal with tooth wear like this by flipping the wheel on its arbour. Am I right in presuming that this is not a candidate for this approach, give the fact that it carries the click arrangement?


As this is a 30-hour clock that sits on display in the local Freemason's lodge (the original purchaser) I imagine it will not be run much. So if there is some life left in those teeth than I'll leave them as they are.

Simon

 

[wow]

Ok, Simon, I have to ask. If the clock is 150 years old and the present owner is the original owner/purchaser, how old is this guy?

 

[Simon Holt]

It's the lodge that's the owner...

 

[R. Croswell]

.......Before cleaning, the clock runs if the spring is more than half-wound. But does that wheel look close to causing terminal failure? Is it a suitable candidate for planishing? I've seen that some people deal with tooth wear like this by flipping the wheel on its arbour. Am I right in presuming that this is not a candidate for this approach, give the fact that it carries the click arrangement?

Simon

In my opinion this wheel is not a good candidate for planishing, there is too much metal that's gone. By the time you you achieve the correct tooth profile the teeth would be too thin to be reliable over time, plus it would take a lot of work. However this wheel is a good candidate for flipping the wheel over. You will need to remove the click and click spring and remount them on the other side, and reattach the wheel after flipping but then it should be good for another 100 years. There is no significant shortening of the teeth so the back side of the teeth look perfect.

RC

 

[Simon Holt]

However this wheel is a good candidate for flipping the wheel over.

Thanks for the suggestion, RC. That would be a first for me. I have a lathe so I can turn up a new rivet. What are the risks associated with driving the wheel off its arbor then driving it back on the other face up?

Simon

 

[R. Croswell]

Thanks for the suggestion, RC. That would be a first for me. I have a lathe so I can turn up a new rivet. What are the risks associated with driving the wheel off its arbor then driving it back on the other face up?

Simon

If you post a picture of the back side of the wheel we can be more specific. Generally the hub extends through the wheel slightly and there is a cup spring washer (brass or steel) that's slipped over the hub extension and staked in place tight enough to hold the assembly together while allowing the ratchet wheel (the hub) to turn as the spring is wound. To get the wheel off you can chuck it in your lathe and carefully turn down the side of the hub extension to remove the disrupted material from staking just enough to allow the cup washer and wheel to be separated from the hub. You won't need a lot of force to separate the parts or pot them back.

Caution - don't reduce the length of the hub extension through the cup washer and don't reduce the diameter of the hub extension, just turn away where it was staked. You will need enough material to stake it back together after the wheel is flipped. Obviously you will mount the click and click spring while the wheel is off. Not too heavy with the staking - keep checking how tight the wheel is. You don't want the wheel to wobble, but if yoiu go too heavy it will bee too tight and hard to wind.

RC

 

[Simon Holt]

I see what you're getting at. Here's the flip-side, RC:

Now I look at it, I'm more nervous about reversing the click spring and its retainer than I am about flipping the wheel and re-riveting the click lever. If it was mine, I'd give it a try. But it's a big responsibility for a relative beginner like me, working on a clock of this age with such significance for its owner. While I would love to give it another 100 years of life, I'd hate to ruin it!

Simon

 

[wow]

It's the lodge that's the owner...

Ok. LOL

 

[Uhralt]

I would flip this wheel over too. Also take a close look at the meshing of the wheel. Based on the wear pattern, it appears that meshing was somewhat shallow. May it would be good to bush T2 and make meshing a bit deeper. It is also possible that the pivot hole of T2 is too wide or worn.

Uhralt

 

[shutterbug]

I would also look at the spring measurements to be sure someone didn't switch it out for a stronger one. That main wheel has taken an unusual beating for a 30 hour clock.

 

[R. Croswell]

To remove the wheel you just turn off the four places where it was staked (3 visible and marked in red) and the washer should slip off without a fight. I would replace the straight brass click rivet with a steel shoulder rivit similar to the one shown here. You may need to clip off the end of the click spring where it is riveted in, but there should be enough left to form a new end when reinstalling. This isn't a difficult repair and there has to be a first time unless you farm out the repair.

I've seen wear like this on other 30 hour movement main wheels. Do check the spring size and let us know what you find and any sign that it has been replaced. Does it look the same as the other main spring? This clock may be old enough to originally have what is generally referred to wrought iron springs. Those have a somewhat rough unfinished look compared to newer springs that have a very smooth surface of newer springs. New springs of the same size as this old style are usually more powerful.

I agree with the others, there may be a depthing (meshing) issue here but I would cautious about attempting to set a deeper mesh. A clock this old likely needs bushings, unless someone has already bushed it. I would begin by installing bushings in all the worn pivot holes and reevaluate the depth of the meshing before trying to change it. If the clock has been bushed at the 1st. and/or 2nd. wheel, a misplaced bushing could also be the problem. Otherwise, its taken 150 years to get this way, and I believe you said that it probably wont be wound very often, so if it runs OK after you flip the wheel I wouldn't worry too much about the depthing.

WARNING: Do not bush the 1st or 2nd wheel pivot holes without first flipping the that main wheel! Those worn "notches" or "craters" or whatever one wants to call them in the wheel teeth now "fit" the pinions as they are in their worn pivot holes. If you install bushings, that will force the main wheel and 2nd wheel closer and the trundles of the 2nd. wheel pinion can "lock up" in the bottom of those notches.

RC

 

[Simon Holt]

Thanks to all. The T2 pivot has a lot of slop, mainly due to it being 'mushroomed' from 1.8 mm down to 1.65 mm:

The mainspring is spring steel, not cast, and 9.5 mm wide x 0.35 mm thick (0.14 inch). How do I find out whether it's an over-strong replacement?

There's another task associated with flipping that wheel that has only just occurred to me – I'd have to remove and re-make the arbour hook in order to get the wheel off the arbor:


At the moment, the clock is cleaned and the time side re-installed. I've given it a full wind and will see how long it runs for before deciding whether to flip that wheel. If it doesn't actually jam in 30 hours, I'll give the owner the choice of whether they trust me to go the extra mile. In the meantime I'm working on making a new alarm stop plate - the thing that goes over the alarm spring arbor and acts like a Geneva stop.

Such fun.

Simon

 

[Uhralt]

There's another task associated with flipping that wheel that has only just occurred to me – I'd have to remove and re-make the arbour hook in order to get the wheel off the arbor:

I think that won't be necessary. The diameter of the wheel opening will be wide enough so that the hook can pass through. Remember that for getting the wheel off you remove the four stake marks on the collet, then the washer and the wheel can come off. The collet can stay where it is.

Uhralt

 

[shutterbug]

There's another task associated with flipping that wheel that has only just occurred to me – I'd have to remove and re-make the arbour hook in order to get the wheel off the arbor:

I would think it would slip off without needing that. I'd be surprised if it wouldn't.

 

[R. Croswell]

The T2 pivot has a lot of slop, mainly due to it being 'mushroomed' from 1.8 mm down to 1.65 mm:
The mainspring is..... 9.5 mm wide x 0.35 mm thick (0.14 inch). How do I find out whether it's an over-strong replacement?
There's another task associated with flipping that wheel that has only just occurred to me – I'd have to remove and re-make the arbour hook in order to get the wheel off the arbor:

Simon

Your 0.35mm spring is approximately 0.014 inch thick and is reasonable for this clock - not over powered. I think you will find that the hole in the wheel is large enough to allow the arbor hook to pass. The excessive wear at the T2 pivot is a contributing factor to the main wheel tooth wear. I would not run the clock in this condition. That main wheel is usable now if you flip it. If you continue to run it the wear will progress to a point where that wheel will become unusable. The second wheel pivot needs to either be turned true and polished or replaced and the pivot hole bushed and the main wheel flipped. There are probably other pivot holes that need bushing as well. I would either do the complete job or set it aside until that is possible.

RC

 

[Royce]

Should you decide to flip the main wheel and relocate the click and click spring, which I believe you will find is not very difficult, I would recommend that when the wheel is off the arbor, do some file work on the ratchet teeth as there appears to me that a couple of the teeth are suspect in their current condition. It is easy to do with the main wheel off the arbor.

 

[Simon Holt]

The diameter of the wheel opening will be wide enough so that the hook can pass through.

I would think it would slip off without needing that. I'd be surprised if it wouldn't.

Doh - of course. I hadn't sussed that the bit with the staking marks is firmly fitted to the arbor, not part of the wheel. Rookie error.

I would either do the complete job or set it aside until that is possible.

Sound advice, RC. I will need the owner to 'buy in' to the risk of letting a novice like me loose on it, versus the risk of running it without the work being done. Personally I relish the challenge of doing something I've never done before, but it's not my clock!

I would recommend that when the wheel is off the arbor, do some file work on the ratchet teeth

Thanks for the suggestion Royce. I was so focused on the main issue that I may well have overlooked that.

Stay tuned for updates!

Simon

 

[Simon Holt]

Well, that wasn't as scary as I thought:

Removing the click components and turning up a new shoulder rivet is tomorrow's job.

In the meantime I have another query. The Geneva stop on the alarm spring is unlike any I've come across before (I've only done two American alarm clocks of this vintage, and they both had the familiar pair of star wheels.) I've seen a photo of what it should look like on this particular clock, and I've made a new stop plate that looks similar:

In operation, it allows 2.5 turns in either direction. Is that how it's meant to work? It's going to limit BOTH how much it can be wound, and how long the alarm will sound for, which makes sense if you pre-wind the spring before fitting the stop plate.

Simon

 

[Uhralt]

In operation, it allows 2.5 turns in either direction. Is that how it's meant to work? It's going to limit BOTH how much it can be wound, and how long the alarm will sound for, which makes sense if you pre-wind the spring before fitting the stop plate.

I think that's the way it is supposed tp work. The alarm will wake the dead, so limiting its duration makes sense.

Uhralt

 

[R. Croswell]

Yes, the stop limits both how much you can wind and how much it will unwind. Fully wound (without the stop) the alarm is very rapid and it slows as the spring runs down. Wind the spring a little less than tightly wound and replace the stop and it should be fine. The alarm will probably never be used anyway.

RC

 

[shutterbug]

I assume that those things were never intended as an alarm to wake you up. Probably more for cooking timers, and things like that. A short duration signal would be enough.

 

[R. Croswell]

I assume that those things were never intended as an alarm to wake you up. Probably more for cooking timers, and things like that. A short duration signal would be enough.

Most of the ones I've seen didn't have a stop, but those that do seem to alarm for an adequate time. I think they might be better as a wake up alarm than a cooking timer for one who gets up the same time each day. I find them difficult to set with any degree of accuracy. I wonder how may of these alarms were ever really used. I don't recall ever servicing one that had any significant wear. I suppose it was at least an extra feature that would justify a higher retail price for the clock.'

RC

 

[Simon Holt]

All your comments about the Geneva stop for the alarm make perfect sense and confirm what I guessed was the case. I'm still curious as to this particular design; the two other examples I've worked on use a pair of star wheels, but I didn't take note of how many turns that design allowed.

Anyways, back to flipping my wheel: I've de-mounted and re-mounted the click arrangement, and all I need to do is stake the wheel back on. I have a problem though: when I turned down the original staking, I didn't realise the collet was not perfectly centred. As a result I've taken a bite out of the collet and can only re-stake in three locations, not four:

The green arrows show where I intend to stake; the red arrow shows where I cannot. I'm thinking about milling a 1 or 2 mm hole into the collet for a brass pin which I can then peen over the spring washer.

Is there a more elegant solution?

Simon

 

[Bruce Alexander]

Hi Simon,

If there is enough material, you might try to deepen the "pocket" or flange so there's a little more brass to peen/stake over the Great Wheel.
Does that make sense? I think you may be referring to it as a collet. Can you deepen that slightly?

 

[Simon Holt]

Hi Simon,

If there is enough material, you might try to deepen the "pocket" or flange so there's a little more brass to peen/stake over the Great Wheel.
Does that make sense? I think you may be referring to it as a collet. Can you deepen that slightly?

Sorry Bruce, I can't picture what you're suggesting. Sorry if I've confused you with my rookie terminology. By "collet" I mean the brass part that includes the click teeth and that is firmly attached to the arbor, as seen in post #18.

Simon

 

[Uhralt]

Sorry Bruce, I can't picture what you're suggesting. Sorry if I've confused you with my rookie terminology. By "collet" I mean the brass part that includes the click teeth and that is firmly attached to the arbor, as seen in post #18.

Simon

I think Bruce is suggesting to make the step on the collet, on which the wheel rides, a bit deeper so that more of the collet will stick out. This will give you more material for staking.

Uhralt

 

[Bruce Alexander]

Yes Uhrait, exactly. Thank you.

I learned that approach from Jerry Kieffer. He referred to the process, if I recall correctly, as deepening the pocket in cases where insufficient depth had been machined in the Factory.

That's not the case here, but you shouldn't need to remove too much material because you were careful not to remove excessive amounts in the first place. You would need just enough to replicate what was there before you started your work. You certainly don't want to compromise the strength of ratchet teeth. We're probably talking about a few thousandths?

A lot of good suggestions in this thread! Perhaps folks who have flipped a number of these Gears will have a better solution that doesn't involve removing more brass. The union has to be rock solid of course.

Although I like to planish gear teeth, I agree that you've taken the right approach here. The Tooth profiles were far too worn for anything other than flipping or replacing the gear. After your work it should be good to go for at least another Century. Probably much more since it's mostly a static decorative display piece now. I would suspect that it will run on less torque from the mainspring too.

Nice job on fabricating the missing Geneva Stop gear btw.

Bruce

 

[Simon Holt]

That sounds good to me. Not something I'd have come up with myself, for sure.

I'll let you know how it goes.

Simon

 

[kinsler33]

They'll come after you when that reversed T1 wears out in 150 years. What will you do then?

I've heard these alarm/kitchen clocks referred to as bread timers, which might have been their actual use. Our collectors will know for sure, but it's my impression that real alarm clocks didn't come into use until around 1900 or so as people began working in factories rather than farms. My 1897 Sears reprint has a couple of fairly standard-looking all-metal alarm clocks, but none of their shelf clocks come with alarms.

The best use I've discovered for these alarms is to set them to go off when the unsuspecting spouse is likely to be sitting quietly reading her usual blood-and-thunder historical novel.

 

[Simon Holt]

They'll come after you when that reversed T1 wears out in 150 years. What will you do then?

Whatever the horological equivalent of Witness Protection is...

 

[shutterbug]

The purpose of staking is to hold the wheel on. As long as that happens, three punches is as good as four.

For setting the timer, you just line up the desired alarm time with the hour hand. It sill alarm within 20 minutes of where you want it.

 

[Simon Holt]

As long as that happens, three punches is as good as four.

That's how I ended up doing it. Due to the eccentricity of the hub, combined with the chamfered profile of the click teeth, I would not have been confident in cutting the 'pocket' any deeper.

The clock is now running on its test stand. I turned back the mushroomed pivot on T2, and with a centre finder it was obvious where the new bushing needed to be.

I had to leave the alarm spring half-unwound in order to get a sensible speed from the alarm train; fully wound, it sounded like an M16 on full-auto... I think Mark's suggestion of it being a bread timer is highly likely. What else would you want to time where 20 minutes either way didn't really matter? High tide?

Thanks to all who contributed. It's given me a lot more confidence in dealing with something this old and this worn. Just think - this clock was made when Billy the Kid was still just a kid, and before the Colt Peacemaker entered production.

Simon

 

[R. Croswell]

...........Due to the eccentricity of the hub, .........

Simon

Was the hub truly eccentric, or was it just not running true in your lathe? I find that main wheel arbors seldom run perfectly true in a 3-jaw scroll chuck. With a 4-jaw independent jaw chuck and dial indicator one can mount a part like this to run with little to no runout. The wear pattern on that main wheel does not suggest that the hub is eccentric. Glad you have the clock back and running. That wheel should last another 150 years, not sure about the rest of the clock.

RC

 

[wow]

Most of the ones I've seen didn't have a stop, but those that do seem to alarm for an adequate time. I think they might be better as a wake up alarm than a cooking timer for one who gets up the same time each day. I find them difficult to set with any degree of accuracy. I wonder how may of these alarms were ever really used. I don't recall ever servicing one that had any significant wear. I suppose it was at least an extra feature that would justify a higher retail price for the clock.'

RC

Maybe that was the first “snooze” alarm. Just wind it back up and turn the little setting dial a hair and go back to sleep.

 

[Bruce Alexander]

What else would you want to time where 20 minutes

I think that a 40 minute margin of error could make for some pretty bad bread! On the other hand, temperature control of the Oven may not have been that precise either. It was probably best just to keep an eye on it. In that regard one of these Alarms would let you go off for while and then come back to keep an eye on things. I've often wondered if the alarm is what turned these into "Kitchen Clocks".

 

[kinsler33]

When Simon's clock was new everyone made their bread in a coal-fired oven. And many if not most ships still used sails.

 

[Bruce Alexander]

Science and Technology, eh Mark? Will it save us from Climate Change? Time itself will still be around even if we aren't puttering around generating electricity by burning coal, throwing away spent toxic batteries or winding up our little minute counting machines.

 

[kinsler33]

Science and Technology, eh Mark? Will it save us from Climate Change? Time itself will still be around even if we aren't puttering around generating electricity by burning coal, throwing away spent toxic batteries or winding up our little minute counting machines.

Clearly no other species worries about time as much as we do.

I'm pretty sure we'll get the environment cleaned up as we learn all the issues and develop remedies.

 

[Bruce Alexander]

I'm pretty sure we'll get the environment cleaned up as we learn all the issues and develop remedies.

You've always struck me as an upbeat optimist Mark.
We (the wife and I) hope that you're right.

 

[Simon Holt]

Was the hub truly eccentric, or was it just not running true in your lathe? I find that main wheel arbors seldom run perfectly true in a 3-jaw scroll chuck. With a 4-jaw independent jaw chuck and dial indicator one can mount a part like this to run with little to no runout. The wear pattern on that main wheel does not suggest that the hub is eccentric. Glad you have the clock back and running. That wheel should last another 150 years, not sure about the rest of the clock.

RC

It was a shame that the arbor was an odd size that couldn't be held in any of my collets. I'd be tempted to acquire a 4-jaw chuck but I'm not sure if one is available for my lathe. I made a poor decision when I bought my lathe - very few accessories are available. At the time, I thought Sherlines were not sold in the UK, and now I know I was wrong. I'm getting close to ditching it but financially it's a big decision.

Anybody out there want a cheap Proxxon lathe?

Simon

 

[kinsler33]

It was a shame that the arbor was an odd size that couldn't be held in any of my collets. I'd be tempted to acquire a 4-jaw chuck but I'm not sure if one is available for my lathe. I made a poor decision when I bought my lathe - very few accessories are available. At the time, I thought Sherlines were not sold in the UK, and now I know I was wrong. I'm getting close to ditching it but financially it's a big decision.

Anybody out there want a cheap Proxxon lathe?

Simon

You can probably manufacture an adapter--I imagine that it would mount on a face-plate, or be a face-plate itself--that'll enable you to mount a smallish Indian 4-jaw chuck on your machine. I've been threatening to do the same for my weird Grizzly 3-in-1 lathe-mill-drill since I bought the thing 23 years ago.

One suggestion: find a metals scrapyard that will let you dig through the debris. It's amazing what you can find, and you might well find a steel disk or something that, with some work, will serve as an adapter. The chucks themselves just bolt on. The scrap yard will just charge you by the pound. Caution: you can load yourself up with fatal amounts of useful-looking junk in taking advice like this.

 

[kinsler33]

You've always struck me as an upbeat optimist Mark.
We (the wife and I) hope that you're right.

We are a clever and adaptable breed. Note the lack of despair amongst our young people. They, and we, will figure it out.

 

[Bruce Alexander]

We are a clever and adaptable breed. Note the lack of despair amongst our young people. They, and we, will figure it out.

Well, to put things back on topic, The clock is ticking.

 

[Simon Holt]

It ran for 34 hours on its first test run. Not bad for a 30-hour clock?

 

[R. Croswell]

It ran for 34 hours on its first test run. Not bad for a 30-hour clock?

Under normal use, if the clock is run every day, it will be wound every 24 hours, so a 34 hour run is pretty good. I'm sure the lodge will be pleased to have their treasured clock working again.

RC

 

[Simon Holt]

Second test run (this time, after oiling the mainspring ) – 48 and three-quarter hours!

 

[Royce]

Excellent job repairing this clock!!

 

[Bruce Alexander]

Second test run (this time, after oiling the mainspring ) – 48 and three-quarter hours!

Simon, that's excellent any way one cuts it. I believe that we have a few 30-hour weight and spring powered clocks in our collection but I've yet to service any of them. Too much to do and not enough time with my current rate of utilization.

You must have really reduced friction and restored tooth form as well as gear depth to the design's ideal specs. Your work as far as to the necessary routine servicing as well as the fabrication of a missing part is first rate in my opinion.

As RC points out your Clients should be very pleased with your efforts. If not, I think that perhaps there's something else wrong in that particular Lodge that you can't fix no matter how hard your try.

I'm curious from a technical point of view. Mainspring torque curves being what they are, how long after a full winding does this clock keep good time?

Regards,

Bruce

 

[kinsler33]

I was going to suggest that if you keep running it you might turn it into an 8-day clock, but that would be confusing so I won't.

 

[Simon Holt]

I'm curious from a technical point of view. Mainspring torque curves being what they are, how long after a full winding does this clock keep good time?

I'll let you know Bruce. I have a camera pointed at it, recording continuously, so I'll be able to plot the curve without needing to stay up all night watching it!

Simon