32: A chipped impulse jewel and another tale of woe

22 02 2018

In post number 30, I described how I made a new escape wheel for a Hamilton Model 21 chronometer and alluded to several other problems to be solved. In this post I will give a brief account of them, followed by a description of how I managed to avoid sourcing and buying a new impulse jewel. As the latter is tiny and rather difficult to photograph without sophisticated equipment, I apologise in advance for the sometimes poor quality of the photos, though I think they convey the details they are intended to convey.

As the fusee is one of the parts first to be removed, it revealed why I believe that the instrument had been dropped while out of its bowl (Figure 1).

Copy of Fusee damage

Figure 1: Damaged chain track of fusee.

Near the top of the fusee, the walls of the chain track had been bent out of straight so that the chain was wedged in place. I gently eased the chain out of the track and then, using a small screw driver with all sharp edges stoned off, I persuaded the walls upright again and filed off tiny burrs with a fine Swiss file. The damage is no longer apparent.

All of the pivots of the escape and balance wheels had been sheared off and the state of their hole jewels attested to the violence of the fall that had broken them (Figure 2).

Holestone damage

Figure 2: Destroyed hole stone of balance.

The plate jewels in the M21 chronometer are friction fitted into the plates and the throat of a staking set of normal size with the usual fitment for pressing jewels into place is generally too small to allow it to be used for a chronometer, so improvisation is needed to achieve a controlled pressure. I fitted the end stones first and then by careful and repeated measurement tapped the hole jewels into place with a brass drift until they were a scant 0.02 mm from touching the end stones. The official manual writes that they should be pressed into contact with the end stone, but it seemed to me that this might prevent the formation of an oil reservoir (if I am wrong, professionals please feel free to make a kindly worded comment). In retrospect I should probably have used a drift held in the tail stock of my small lathe as an improvised press with the plate resting on the  face plate.

I have described pivot repairs in posts numbers 6 and 7.

In early M21 chronometers, the plates and the detent were made of nickel silver, a brass in which half of the zinc has been replaced by nickel to improve its mechanical and corrosion resistance properties. There is no silver in nickel silver, but it is a shiny white metal. Columbia Metals UK describes it thus: “Nickel silvers are capable of providing a unique combination of strength, high modulus spring properties, corrosion and oxidation resistance, …. and numerous other attributes combined with ease of forming, machining, plating and joining.” Later chronometers have detents of beryllium copper. Happily, in the disaster, which seems to have included running away of the movement, my detent was “bowed but not broken” and the passing spring had survived . The two photos merged into one of Figure 3 were taken on different days in obviously different lighting, but are of the same detent from different viewpoints.

Detent merged

Figure 3: Bent detent.

The top image shows the detent in its support block and the bend at the base of the spring is rather obvious. Less obvious in the lower image is that the bend is not symmetrical and has introduced a slight twist. Only a fragment of the locking jewel had survived. By this stage, I had removed the passing spring. A few words of warning may be useful here. The screw that secures the spring in place is the tiniest part in the whole chronometer and can easily be treated as  a fragment of dirt and be lost, so it pays to store it safely in its own little container.

I straightened the spring by drawing it between the rounded jaws of special pliers, as described in post number 28 and eventually restored it to straightness. In an effort to avoid expense which I can ill afford, I attempted to make a new locking stone of tungsten carbide, as described in post number 24, for a Soviet MX6 chronometer, but the M21 locking stone is rather more slender at only 0.6 mm in diameter (versus 0.8 mm) and vibration from the belt of my tool and cutter grinder (since remedied) was enough to break it off when I attempted to grind the flats. I was thus obliged to buy a new locking stone and fit it as described in post number 15.

In early M21s the rollers were in one piece so that the mutual angles of the unlocking and impulse jewels was fixed, as in my example. Quite early on, the one-piece rollers were replaced by two separate parts. Figure 4 shows a view from a larger photo of the rollers from above, taken so I could replace them in their original position.

Copy of Rollers

Figure 4: Roller and jewels.

At first sight, all appears well, but as repairs progressed I discovered that the impulse jewel was quite badly chipped, so much so that a “replace and hope for the best” attitude would have been a poor plan. Figure 5 shows the jewel after removal.

17 Damaged impulse (2)

Figure 5: Chipped impulse jewel.

As I had not at that stage found a source of a new jewel, and doubted my ability to make a new one from tungsten carbide, I decided to reface it using a diamond file and laps, but first I had to devise a method of holding this tiny piece of synthetic ruby.  I filed a notch into a scrap of 2 mm brass sheet using a warding file until the jewel fitted snugly, and then secured it in place by melting a flake of shellac over it (Figure 6).

18 Impulse j set

Figure 6: Impulse jewel set into holder.

Note that this view shows that the damage extends quite a way down one edge of the jewel, so that in filing the new clearance the length of the jewel was quite considerably reduced. Looking through that most valuable of instruments, the retrospectoscope, I should have seated the chipped side in the holder and filed the new clearance on the other end. This would have maintained most of the length and made for a more secure re-fitting in the roller. A diamond file cut away the damaged face, much more easily than the same file cuts into tungsten carbide, and this underlines how much harder the latter is than ruby. As the clearance face comes into contact only with air, there was no need to achieve a very fine finish, but rather than leave a sharp edge, I rounded it a little with fine laps.

The roller is 6.35 mm (0.25 inch) in diameter, so another scrap of sheet brass with half of a 6.3 mm hole filed away provided the gauge to ensure that the tip of the jewel projected the correct length when being secured in place with melted flake shellac. This is described in post number 17.

Figure 7 shows the final position. Although the hold of the roller on the impulse jewel looks precarious, the shellac covers the base of the stone and I have left fillets of shellac on either face. If it does eventually give way, I will replace it with a carbide version.

20 Impulse j reset

Figure 7: Impulse jewel refitted.

The rest of the overhaul was routine and the chronometer showed no reluctance to resume its ticking with a very good action. However, I will end with a cautionary tale.

After about a week of rating it and finding that it had a very steady losing rate of about 1.7 seconds per day, its rate became erratic and then it lost two seconds overnight, so I stopped it and began to look for a source of the problem. As I was checking the balance and escape wheels for excessive end or side shake, the instrument suddenly declined to run at all. Suspecting that the impulse jewel had perhaps shifted, I removed the balance to check and found that it was still secure so I replaced the balance and turned my attention to the detent, to check what I already knew to be the case, that the escape wheel teeth were engaging about one third of the acting face of the locking jewel. The engagement was now much deeper. It was at this point that I discovered that, having at original assembly made a slight tweak to the depthing of the passing spring, I had failed to re-tighten the screw that holds the detent to the mounting block, labelled “detent clamp screw” in the original manual. Once I had tightened it, normal running resumed. Now I’m investigating its rate when half wound.





31: An eight day chronometer

14 02 2018

Some weeks ago, I received an e-mail asking about an eight day chronometer that the writer had recently inherited. He was able to tell me that it carried the name of John Carter of Cornhill, London and had been sold to the Admiralty in 1872. Its history between then and 1936 was not known, but in that year, it was sold by the Admiralty and in 1946 it had travelled to India with Ron’s uncle where it had remained on a bookshelf in a hot and humid climate until late in 2017.

After an exchange of e-mails, Ron sent me some photographs, which he has agreed I could share with others in my blog. Most of the photos can be enlarged by clicking on them. From Figure 1,  we can see that Carter did not actually make the movement. It was very common for retailers and even other makers to buy in finished movements and put their own names on the face. Makers would also buy from other makers if they had an order that they could not meet at the time. Parkinson and Frodsham as a company were active between 1801 and 1890. I cannot see a number anywhere on the movement that might give a clue to date of manufacture.

7 John Carter 738

Figure 1: Top plate inscription.

The form, size and layout of movements changed remarkably little from about 1800 until manufacture ceased in the 1980s and the same can be said for the bowls and cases. Two day chronometers of whatever make usually had three-tier cases forming a cube with sides of  185 mm, and heavy brass bowls of about 105 mm diameter. Almost always, the cases were brass bound with brass corners, presumably to hold them together in the marine environment if the glue and pins that held them gave way. Curiously, sextant cases, which would be a good deal more exposed to the elements than chronometer’s were almost never brass bound and the only one I have ever seen that was, other than so-called “reproductions”, is in my possession. Straight away, we see from Figures 2 and three, that my correspondent’s chronometer is an exception in the other direction.

1 john carter 738

Figure 2: Front of case.

2 John Carter 738

Figure 3: Side of case.

The front of the case has an interesting little escutcheon around the key hole, a bone or ivory lozenge, perhaps for the Admiralty to add their accession number to, and a broken button catch for the top lid. It is quite hard to find replacements for these little catches. The side view shows that the corners have rebate joints, another good reason for having the case brass bound, as they are less secure than dovetail or comb joints. We also see the two screws that secure the gimbals lock, the screw and rectangular washer that holds one side of the gimbals bearing and a standard drawer pull type of handle.

3 John Carter 738

Figure 4: Bowl in case.

On opening the case, we see a tipsy key  and gimbals lock whose forms did not change over time and between different makers of various nationalities. Unusual, however, is the way the tip of the gimbals lock engages with a brass band that nearly encircles the bowl, whereas usually it engages with a slot in a simple bracket that is attached directly to the bowl. The 15 mm thick walls of the case have dust seals of a darker wood, possibly ebony, let into the top of the bottom tier, except where the box lock and the hinges intervene.

4 John Carter 738

Figure 5: Face.

In an otherwise unremarkable 19th and early 20th century face is a clue that reveals the main interest of this chronometer (which is not to say that this venerable instrument is not desirable for other reasons). The state of wind indicator shows that the chronometer has a movement that runs for eight days. It is in a fully-wound state, while the minute hand is out of step with the seconds hand, whereas it should indicate a full minute when the seconds hand shows 60 (or zero, which amounts to the same thing).

Eight day chronometers are very uncommon and it is a little difficult to understand why any were made, since they would in practice be wound daily at the same time of the day to maintain a constant rate. The famous  and influential Captain Lecky, in his “Wrinkles in Practical Navigation“(ninth edition, London, 1884), writes “…it has been found that eight-day chronometers do not preserve altogether the same rate throughout the week; that is to say, that (though other conditions may be the same) their daily rate towards the end of the week will not agree with their daily rate at the commencement of it; notwithstanding which, the mean  rates of two consecutive weeks  may agree exactly.” He also adds “On account also of the lightness of the balance, the eight day chronometers do not go so well on board steamers which suffer much vibration from their machinery.”

Rupert Gould in his authoritative “The Marine Chronometer” confirms this when he writes (p.218) “...there is absolutely no advantage gained by making a chronometer go for more than two days between windings, and such machines are inferior both in principle and detail to the ordinary two-day pattern, although, if well-made, they may be found quite satisfactory in use at sea.

I reassured my new internet friend that it would not hurt to unscrew the bezel and tip out the movement. Figure 6 shows more details of the exterior of the bowl and the strange provision made for the tongue of the gimbals lock and the dust cover for the key hole. Note that the disc that houses it, attached with three screws, has the letters “H.S.”, followed by a broad arrow indicating government property, and the letter “I (or number 1). Perhaps we may tentatively wonder whether these letters refer to the hydrographic service of one of Britain’s former colonies.

6 John Carter 738

Figure 6: Exterior of bowl and GA of movement.

Turning to the general view of the movement, the most obvious difference from a two day chronometer is its exceptional height, necessary to accommodate a taller spring barrel and fusee, and that the rest of the movement, except for the centre wheel arbor, is sited between a sub-plate and the bottom plate. Figure 7 shows some more details of the spring barrel and fusee.

9 John Carter 738

Figure 7: The power source.

The the tall, large-diameter spring barrel accommodates about eight turns of  chain, while the fusee has about 16 turns, nearly twice as many as a two day chronometer, while the great or fusee wheel seems to have many more than the standard 90 teeth.

Another view of the power source (Figure 8) shows at lower left the elongated click of the maintaining power ratchet wheel attached to one of the pillars and, at top right beneath the top plate, the short pillar for  the fusee iron. This latter proved rather hard to photograph, so a verbal description will have to do. The fusee iron is shaped somewhat like a fork with two prongs. Its base is hinged to the short pillar and held off the top plate by a leaf spring. As the chain mounts the fusee during winding, it comes into contact with the underside of the fusee iron and raises it parallel to the top plate At about this point, a projection, called the snail, which is screwed on to the top of the fusee, enters the fork and brings winding to a halt. In the absence of  stop work, there is nothing to stop the heavy-handed from breaking the chain, with often disastrous consequences for the movement when the large amount of energy in the oversized spring is suddenly released. Also in Figure 8, the very large wheel for the state of wind indicator is seen at bottom right.

11 John Carter 738

Figure 8: Maintaining power.

This leaves the escapement, shown below in Figure 9. I have been able to show only the balance wheel and its associated parts The bi-metallic rim of the wheel is of course split adjacent to each timing weight which are situated as usual at the ends of the spokes, while large, wedge-shaped weights for temperature compensation are at roughly right angles to the spokes. The upper end stone seems to be of diamond mounted in a two-part balance cock. It is just possible to see the ruby locking jewel to the right of the spoke at seven o’clock and the ruby end stone of the escape wheel arbor to the left of the spoke. While in most modern chronometers the upper balance spring collet was attached to the top of the cock, in this instrument it has its own little pillar, and I wonder whether this might have been some form of compensation for middle temperature error, if the horizontal strip turns out on further examination to be bi-metallic.

Copy of 10 John Carter 738

Figure 9: Balance wheel.

I hope that this account may be of interest to those who, like me, have never seen an eight day chronometer “in the metal” as it were. Many more details of the structure of chronometers may be found in my book “The Mariner’s Chronometer”, available via amazon.com




30: A new escape wheel for M21 chronometer

6 02 2018

I recently acquired a damaged chronometer that seemed to have been dropped on to a hard surface while out of its bowl, or perhaps the owner had thrown it at a wall in frustration. At any rate, the walls of the fusee track had been squashed in at one point and had trapped the chain. By carefully using a small screwdriver whose edges had been rounded off, I was able to open out the track and a fine file removed all other traces of the injury to the fusee. On further inspection I found a large range of problems that would need to be fixed before the chronometer could be brought back to life, some of them probably due to the shock of the fall and others probably consequent on the initial damage.

The most obvious was that all the pivots of the balance and escape wheels were broken, which I could fix, and two of the escapement hole jewels were shattered, which I could not, so I was committed to having to buy replacement jewels. In the Hamilton M21 chronometer, these jewels are friction fitted in the holes in the plates and so are rather more difficult to fit than those of most other chronometers. The locking jewel was broken off flush with the top of the detent, but happily, enough remained for me to be able to photograph, so that its expensive little replacement could be replaced at the correct angle of 8 to 12 degrees of draw. The detent spring was also bent out of shape, but I was able to straighten it using the method outlined in Post 28.

To the naked eye, the escape wheel looked fine, apart from the disaster to the pivots, but on closer inspection its teeth looked worn and when viewed under a low power microscope, the tips and sides of the teeth looked worn, as shown in Figure 1.

00 Worn teeth

Figure 1: Worn and damaged escape wheel teeth.

Normally, only the tips and front faces of the teeth come into contact with anything, i.e. the locking jewel and the impulse jewel. Its manufactured diameter would have lain between 13.14 and 13.18 mm, but its measured diameter was only 12.73 mm, a relatively enormous disparity, so I guess that something had caused the escape wheel to “run away”,  the while battering itself against a jewel, eventually breaking the locking jewel and, as I found later, badly chipping the impulse jewel. Perhaps, when the first escape wheel pivot broke, perhaps it was still able to run, albeit drunkenly, and so damage the sides of the teeth.

I searched for a new escape wheel, but could not find one to buy. Most chronometer escape wheels seem to have fifteen or thirteen  teeth, but Hamilton chose sixteen, so the readily available Soviet MX6 escape wheel (a bargain at US$ 235 with pinion?) could not be substituted. I had to bite the bullet and try to make one myself. A quick trawl of the internet showed that two people had made escape wheels by laboriously making a punch and a die, so that the wheel could be punched out of sheet metal with the dimensions and crossings all ready formed for machining the teeth. This method did not appeal to me, and anyway, for me part of the pleasure of restoring these instrument comes from using old techniques. A punch and die are fine if you want to make lots, and Hamilton eventually made over 11,000, but for a one-off, it seemed to me to be overkill.

Figure 2 is a posed photograph of how to remove an escape wheel from its arbor.

Figure 2: Removing wheel from arbor.

Note how a slot in a scrap piece of brass protects the collet from damage, while the arbor is punched out. Beginners in the use of a staking tool perhaps need to be reminded to check the depth as well as the diameter of the hole in the punch, so that they do not inadvertently damage the pivot. The photos in what follows were taken while developing a method over several days, so the keen-eyed reader may notice certain lapses in continuity…

Some makers start by cutting out or buying a circle of brass, mounting it on an arbor, turning it to size and then cutting the teeth. This works well with larger wheels, but because of the small size of the hole in the middle, this makes securing the blank a bit uncertain for turning and the set up lacks rigidity when it comes to cutting the teeth. I used the more wasteful but safer method of machining the teeth on brass rod and then parting off slices, so I could have several goes at developing my method. After turning the rod down to the outside diameter of 13.16 mm, I transferred the chuck from the lathe to my dividing head in a bid to retain concentricity. Figure 3 shows the faces of the teeth being cut.

1 Cut front face

Figure 3: Cutting the front of the teeth.

Although the photo shows the work piece in a three jaw chuck, I found that the set up would not hold concentricity between lathe and dividing head to any better than 0.03 mm, so I eventually used a four jaw chuck and centred the piece with a dial indicator against a register machined on the bar. To cut the teeth, I used a fly cutter filed from a piece of 6 mm silver steel and then hardened, tempered and polished on the cutting faces. Run at 1,500 r.p.m this gave an excellent finish that needed little polishing. Tools for cutting brass need to be sharp. Note that the acute angle of the teeth of about 60 degrees means that the front edge of the tool needs to be lowered below centre by just over 2mm.

It was somewhat more difficult to machine the curved back faces of the teeth, but I eventually managed to file a radius of 2.5 mm on the end of a piece of silver steel and to form the all-important relief behind the cutting edge.

2 Cut back face

Figure 3: Back faces of tooth being cut.

Setting the tool height to get the tooth looking correct needs to be combined with rotating the work piece to get a tooth that is robust enough and to leave a narrow land at the top of the tooth, specified as 0.13 to 0.14 mm wide. Of course, this cannot be easily measured and for myself I modified that specification to be “land present and just about visible”. In this photograph of an early attempt, the land is a little too wide, but the form looks fine. If you take too much off, the diameter inevitably has to be reduced to restore the land.

7 Set parting off

Figure 4: Parting off.

Most turners will agree that “Parting off is such sweet sorrow”. The tool needs to be sharp with its face square to its axis, and the axis has to be square to the axis of the lathe. The cutting edge needs to be exactly at centre height or, with a back tool post, a minute amount above. With the work piece transferred back to the lathe begins the task of parting off slices while ensuring that no slice is thinner than 1.32 mm. As the conditions above are difficult to meet exactly, it is better to part off oversize and then face the slices down to the desired thickness. Measuring small distances in a confined space is best not done with a ruler, and I use a depth micrometer, as shown in Figure 4. Parting off is easier if there is a central hole and I finally remembered to drill this before parting off.

I then turned a holding fixture which would not leave the lathe until all the slices had been reduced to the correct thickness, though I eventually realised that Hamilton’s specifications and tolerances were to ensure interchangeability and that 0.1 mm either way was of little importance as long as it is no wider than the impulse roller. Figure 5 shows the simple fixture being bored to produce a recess 1 mm deep that would just accept a slice of embryo escape wheel.

3 Bore fixture

Figure 5: Boring fixture.

Knowing the depth of the recess from the outer shoulder, it was then possible to set a facing tool back from this shoulder, using the graduations on the tool slide, by an amount to give an over all thickness of 1.27 to 1.32 mm.

At first, I used shellac to secure the wheel in place, but found that the fixture combined with the mass of the chuck formed such a large heat sink that it was difficult to reach a high enough temperature with a small flame, a problem not eased by relieving the end of the fixture with a deep slot, seen in some of the following photographs. Eventually, I used superglue. It melts at a much higher temperature than shellac, so I was obliged to remove the fixture from the lathe and soak it in acetone overnight to release the wheel.

4 Counterbore wheel

Figure 6: Initial counterbore.

Having cemented the part into place and faced it to the correct thickness, the counter bore in the wheel can be started using an 8 mm end mill or slot drill, its depth controlled by the graduations on the tail stock quill. This counterbore is then opened out with a boring tool (Figure 7) to its correct depth and diameter.

5 Counterbore enlarge

Figure 7: Counterbore enlarged.

The correct diameter removes just a little of the root of the tooth to give the finish shown in Figure 8. To ensure concentricity of the hole in the centre and the tips of the teeth, I ran a small but rigid reamer, with one of its two end teeth ground back to make of it  a small but rigid tool. A reamer is usually used to size a hole and will normally follow the existing hole, but in this case, only a whisker was removed, and in any case, the hole is too small for a conventional single point boring tool to enter.

6 Counterbore finished

Figure 8: Counterbore completed.

After giving it a good soak in acetone the wheel could be removed from the fixture so that marking out for the crossings could begin. I faced the ends of a piece of wooden dowel and glued to each end a piece of emery paper, one of 800 and one of 1200 grit and rotated the dowel against the floor of the counterbore to remove most of the turning marks prior to marking out. In a scrap of brass, I faced, drilled and reamed a 3 mm hole and made a close fitting removable spigot, one end of which was turned down to a close fit in the hole. Before parting off the spigot I made a minute centre mark in it with a sewing needle held in the tailstock chuck (Figure 9).

8 Marking out 1

Figure 9: Marking out jig.

From this centre I scribed a circle of radius of about 18 mm and divided it into 6 parts by the well-known method of stepping the dividers around it at the same radius. I marked out centres on three of the radii at a radius of 13 mm and from these centres scribed the outlines of the spokes, adjusting the radius by trial and error to give sufficient metal at the joining of the spokes with the periphery, which has a radius of about 5.5 mm. Midway between the spokes and the periphery I made punch marks and then removed the wheel from the jig to drill 2 mm holes at these points. Figure 10 shows a wheel at this stage.

9 Drilled

Figure 10: Ready for crossing out.

The purpose of the holes is to allow entry for the blade of a piercing saw, in this case a new 4/0 blade, but before beginning to saw, some filing makes subsequent sawing and filing much easier. I long ago made myself a mini four square file by grinding away two adjacent sides to make safe edges and I used it to file in and out to the marked lines, so the the saw could start right next to the line (Figure 11).

10 Extend hole

Figure 11: Preliminary filing to lines.

The wheel is much too fragile to be held in a vice and Figure 11 shows how it is held horizontally between fingers and a horizontal surface while the file moves up and down. It helps greatly to be able to see exactly where the file (or saw blade) is going and I have a binocular microscope mounted on a boom on my work bench. Note too the piece of leather between my fingers and the wheel. The teeth of the wheel by this stage are sharp! A later photo (Figure 13) taken before this one shows my fingers before I learned this important lesson. Figure 12 shows a wheel prior to sawing.

11 Drilled and filed

Figure 12: Preliminary filing completed.

Sawing could now commence, again holding the wheel horizontal, as is usual when using a piercing saw (Figure 13). The saw “table” is simply a strip of metal cantilevered from a small vice to give the hand room to move up and down beneath it, and the wheel is rotated to keep the blade tangential to any curve as the cut progresses. When progressing around tight curves or into corners, the blade must be near-vertical, but around shallow curves or in straight lines the blade seems to follow the lines better if canted forwards a little, as shown in the figure.

12 Sawing

Figure 13: Crossing out with saw.

It will, I hope, be obvious after a little thought that only half of each crossing can be sawed this way and for the other half the wheel must be transferred to the other side of the table and the saw held in the left hand. This needs more ambidexterity than I have, so I simply reversed the blade in the frame, so that the teeth faced inwards towards the frame, and sawed backwards towards myself. (See Fergus’s comment) Figure 14 shows the results in an early, practice attempt to assess the practicality of making the crossing by hand.

13 Sawed

Figure 14: Ready to file again.

The rest involves filing to the lines. The closer one can saw to the lines, the less filing is required. Swiss needle files are needed and there is a particular form used for crossing out and it is called a crossing file;  both surfaces are curved with different radii and tapering to a point, so with care, finely rounded internal corners can be cut. A crochet file is useful for getting into sharp corners as it is tapered in width and in length. Again, a powerful aid to vision is very helpful.

A close inspection with a microscope in good light will show all manner of burrs and the simplest way of removing them from corners is to lightly draw the blade of a small craft knife across them, taking great care to avoid the faces and tips of the teeth.

Each face of the wheel is easily polished by rubbing against a piece of wet and dry emery paper resting on a scrap of plate glass under water to which a drop or two of washing up liquid has been added. I start with 800 grit and finish with a piece of well-worn 1200 grit, taking the polishing no further, as these faces contact only air.

The acting faces of the teeth were left with a very fine finish by the fly cutter, but I felt that the locking and impulse jewels would have an even smoother ride if polished (Figure 15).

14 Polish teeth

Figure 15: Polishing acting surface of tooth

The figure shows how I held the wheel, by now with its collet in place. In a block of wood that I could hold comfortably in my fist while it rested on the bench, I let in a piece of pivot steel and held the wheel stationary with an index finger, while polishing the faces under direct vision with a scrap of diamond-impregnated film glued to an old feeler blade. I went from 9 to 3 micron film and decided that was far enough.

Figure 17 shows the original wheel with a couple of trial wheels, with the one on the left nearly good enough, but as the rim at the top is a little irregular, I decided to finish with a spare one of the dozen or so that I had parted off at various stages of my trial. I had fitted this to the chronometer before I thought to photograph it, and as the chronometer has now run for over 24 hours with a gain of 1.7 seconds with an excellent action and the mainspring set up only two turns, I am not about to take it apart for a photograph. The final version is seen in Figure 15.

15 Group finished

Figure 16: A compendium of escape wheels.

If you compare the finish of punched out edges, as shown in Figure 1 with the edges left by filing in Figure 15, there is surprisingly little difference (Figure 17).

21 Teeth compared

Figure 17: New (right) and old (right) wheel finish compared.

My teeth probably have more mass than the originals , but the rim is a little finer and the mass of the spokes is concentrated nearer the centre, so that the inertia, mr², of my wheel is probably about the same.

Making the collet was a simple turning operation, albeit at a small scale and industrial glues have made interference fits and riveting of collets unnecessary. Finally, Figure 17 shows a Soviet MX6 escape wheel being refitted to its arbor.

16 Replace on arbor

18: Fitting an escape wheel to its arbor.


I hope you enjoyed reading this post and if you haven’t already done so, I encourage you to buy my book, available from amazon.com. It may well tell you more than you wish to know about the structure of the marine chronometer.







29: More on Tipsy Keys

3 01 2018

A note on finding your way around this site: If you know what you are looking for, you can enter a search term in the search box. You can also look in the “List of Posts” to get a date and either click on that date or enter a search term.

In my post Number 8 of July 30, 2013, I wrote about how to make one form of tipsy key. Recently, to amuse myself, I have bought shipwrecked chronometers and tried to bring them back to life. Currently, I am working on a Hamilton Model 21 chronometer which seems to have been dropped, probably while out of its bowl, as all the pivots of the balance and escape wheels are broken and the track on the fusee for the chain has been damaged. Two of the hole jewels of the escapement are also badly chipped and the locking jewel has been broken. While waiting for replacement jewels, I have occupied some time by cleaning the rest of the chronometer and making a new key for it. Since a lot of time is spent setting up for machining parts, it is sometimes almost as quick to make two parts as to make one and this is what I did.

0 Two keys

Figure 1: Key fashions.

Figure 1 shows two tipsy keys that use clutches to prevent winding a chronometer clockwise. The one on the left is from an antique Mercer chronometer that perished when the Australian city of Darwin was largely destroyed by Typhoon Tracy in 1974 and the one on the right is from a German Einheits-Chronometer of 1943. The latter still keeps an excellent rate.

Figure 2 is a composite drawing of the interior mechanism of the second key, while subsequent drawings of the separate parts give the dimensions of the parts as I made them. I had to alter some of the dimensions of the knob so as to use what I had available in the way of brass stock, but as long as the square hole fits the winding square of the chronmeter, the dimensions do not have to be identical to the original.

Tipsy key composite drwg
Figure 2: The interior of the key.

The clutch teeth on the end of the winding shaft inside the conical body are held in engagement with the teeth on the knob by means of a short helical spring. The teeth are sloped so as to slip if the knob is turned clockwise but to engage if turned anti-clockwise.

It is a simple turning exercise to make the shaft (Figure 3), by turning down a piece of 8 mm stock to 6 mm, drilling the end to a depth of about 25 mm to a diameter equal to the across-flats dimension of the chronometer winding square, and parting off.

Tipsy key shaft drwg

Figure 3: Drawing of shaft.

The round hole then serves as a guide to convert the round hole into a square one, while the depth of the hole allows a square needle file to get a decent grip on the metal. The nearly completed filing is shown in Figure 4. Of course, those lucky people who possess small broaches could use them and it is also possible to drill square holes with an appropriate attachment to the drilling machine.

New key 003

Figure 4: Converting a round hole to a square one.

Tipsy key body drwg

Figure 5: Body drawing.

For the body, I turned down a piece of 20 mm round brass bar to 18 mm, drilled and reamed the centre hole to 6 mm and followed the 6 mm hole with an 8 mm slot drill to form a flat-bottomed hole as shown in Figure 6. As I was making two keys, for each tool set up I switched ends of a length of bar.

2 Counterbore 8

Figure 6: 8 mm hole counter-bored.

The top slide is then set over to 12.5 degrees and the conical outside generated (Figure 7), before parting off to length. Technically, I suppose the shape should be described as a frustrum of a cone.

2.1 Body turn

Figure 7: Generating conical shape.

I postponed drilling and tapping the M3 (or 6 BA would do) holes until the knob was made. Thirty millimetre  stock would have been better than the 26 mm I used, as it would have given me a key with a bit more leverage.

Knob drrwg

Figure 8: Knob drawing.


2.2 Knob turn

Figure 9: Turning diameters of knob.

Making the knob begins by turning  the 8 and 20 mm diameters shown in Figure 8. The bar is then set up in a vice on the milling machine to form the flat parts of the knob (Figure 10). This is perhaps a good spot to point out that tools for cutting brass need to be sharp and preferably reserved for use only on brass, as once they have been used on steel they tend to skid over brass.

3 Mill flat 1

Figure 10: Milling flats.

The second flat is formed  by rotating the bar through 180 degrees and checking with a micrometer that the surfaces are parallel before making the final cut on the second surface. While it was still attached to the bar, I used a simple template to mark out the complex shape (Figure 11), before cutting around it using a piercing saw. No doubt it could be machined using some sort of computer-aided process, but with a little practice and a sharp saw blade it is much quicker to cut it by hand and finish off with sharp files.


4 Template

Figure 11: Using template to mark out for sawing to shape.

This is perhaps a good moment to drill holes at tapping size for the screws in the knob. They will be used to spot the holes in the body and then enlarged to a clearance size later.

Next comes the machining of the clutch teeth, and it needs some thought and care to make sure that they slope in the correct direction! The knob can be held by the flats in a machine vice that tilts. I guessed that the teeth tilted at 30 degrees, but 25 degrees would have given more clearance for the side of the 3 mm diameter end mill at the end of its cut (Figure 12).

5 Mill knob tooth

Figure 12: Clutch tooth being formed.

The same set-up is used for the teeth on the end of the shaft, as shown in Figure 13.

6 Mill shaft tooth

Figure 13: Milling teeth on shaft.

After cleaning off burrs with a fine file, the key can now have a trial assembly and a spring cut to length so that the teeth are held in engagement and that there is enough free space for them to slip when turned clockwise.

Drilling for the tapping holes in the body presents minor problems, as its shape prevents it from being held in a vice. I got around this by assembling the key, holding the shaft in a drilling vice and rotating the body until the teeth were locked, then drilling the hole to full length. Though brass is traditionally cut dry, it is a sound plan to use a little lubricant when tapping the holes, as there is a tendency to jamming unless the taps are sharp. Most of us cannot afford to keep two sets of taps, one for sole use on brass, so a little lubricant may save a lot of heart ache due to a work piece having to be scrapped because a broken tap is jammed in it.

A little grease and assembly with a couple of countersink-head screws completes the key (Figure 14).


Figure 14: Two new and two old.


I hope this account was of interest to you. You will find much more about marine chronometers in my book. Take a look at “About the Mariner’s Chronometer”.






28: A tale of woe that ended well.

24 11 2017

Apart from Soviet era MX6 chronometers other chronometers are out of my financial reach, unless I buy damaged ones “for parts or repair”. As I now have several MX6 s I can only justify buying another if I challenge myself to right the wrongs it may have suffered. At the beginning of October I returned from the USA with a homeless MX6, of which the seller had said it would not wind beyond 48 hours, whereas 56 hours is the norm. The Department of Homeland Security had rummaged through my baggage, presumably upon seeing with X-rays a dense circular object  with clockwork, and had replaced the layers of bubble wrap after a fashion, so further damage had not occurred.

Once recovered from a 15 hour flight from Houston to New Zealand, followed by a 5 hours drive to my home in the Far North, I set about exploring the innards of the instrument. The end of the chain around the barrel was at the back and  the clock ran when started, so, rather incautiously, I wound the clock to 48 hours and continued, forgetting that the stop-work needs a full wind to operate. There came a loud snapping noise followed by a frenzied whirring…Upon opening the machine I of course found that the barrel end of the chain was no longer attached to the barrel, but of a hook there was no sign (Figure 1 – click on the photo to enlarge and use back arrow to return to text)). The barrel arbor had cast off its ratchet, so happily the whirring had come from the barrel rather than from the movement.

Broken chain labelled

Figure 1: Broken and short chain.

Note how the chain, even if it had had a hook, is still one turn of the fusee short of reaching the stop bar. It seems that someone, not necessarily the seller, had simply stuffed the broken end of the chain  into the barrel slot. In any event, when the chain lost contact with the barrel, either the free end or the recoil of the movement had done a lot of damage. The moral of the story is that you should not wind an obviously defective chronometer to see whether it goes. It should have been obvious to me that a bit of the chain was missing and that the stop work would not operate, risking breakage of the chain at the end of winding if it had a hook into the barrel or, as I found out, the chain simply let go of the barrel. The chain was about 120 mm short of the required length of 850 mm and I replaced it with steel cable as described in post number 27 of 13th February, 2017.

Once fully dismantled, I found that the upper pivot of the escape wheel had been repaired by the classical method of drilling down the broken end and letting in a new pivot. The new pivot was rather short and it had not needed much to knock it out of place. I had a spare escape wheel arbor and pinion which had a broken upper pivot and I used my preferred method of repair, by using a muff, as described in post number 7 of 22nd July, 2013. Using the classical method, if the tiny drill, around 0.6 mm in diameter, breaks off in the broken arbor, you may not be able to get the broken stub out of the hole and, it being made of high speed steel, you won’t be able to drill it out.

That was the easy bit. The detent spring, which must have been intact when I first tried it, because the clock ran, had taken on a Z-shape. At least it was not broken, and the fact that it had distorted without breaking gave me some hope that I might be able to straighten it. While is was possible to straighten a passing spring by drawing it between my finger nails (see post 22 0f 20th June 2016), the detent spring is made of sterner stuff and so I used a pair of pliers with circular jaws, bending the spring while drawing it between the jaws, as shown in Figure 1, which is posed with a slip of brass shim between the jaws.

Straighten spring 001

Figure 2: Technique to straighten spring.

Drawing the spring between the jaws, while angling the pliers against the bend, irons out any kink, so that a reverse bend is not simply added to the original bend. Eventually I managed to get the spring straight again without breaking it.

The next step was to replace the passing spring. While some chronometers had oval holes in the passing spring to allow for some adjustment, in the MX6, there is no provision for this and the tip of the passing spring projects rather less than a millimeter beyond the tip of the horn. This means that in adjusting the depth of the detent, this is the total range of available movement and in lifting the passing spring the discharging jewel must pass clear of the tip of the detent. On the return, the jewel must lift the locking stone off a tooth of the escape wheel far enough to unlock the wheel. If the locking stone is too deep in the escape wheel teeth, unlocking won’t happen, so some adjustment of the banking screw (or stop button in the Hamilton M21 escapement) may be necessary. It seems to be about right when a tooth of the escape wheel overlaps about one third of the width of the jewel face.

In making these interdependent adjustments, I start with the locking stone and check that every tooth of the escape wheel is locked by this amount, just in case the upper pivot is bent or for some reason there has been uneven tooth wear. Once I have done this, I then start with the tip of the detent well clear of the discharging jewel and move it in very gradually, operating the escape wheel with a finger until the passing spring is lifted off the horn of the detent. Only then do I check that unlocking takes place on the return stroke. If the depth of the detent is set too deep, the discharging jewel will strike the back of the horn on the return, instead of the passing spring, and refuse to go further because the banking screw stops it, a very good reason for checking the operation of the escapement by hand, rather than under power, which would risk breaking the discharging jewel or a balance wheel pivot.

Then one can wind the chronometer a turn or so and let it run under power, being prepared to stop it at the first sign of tripping, which is usually cured by increasing the depth by a tiny amount each time. This of course assumes that you have not disturbed the mutual angles between the upper balance wheel spring stud, the discharging roller and the impulse roller, another interdependent set of adjustments. I knew that I had all the angles correct and it needed less than an eighth of a turn of the depth-adjusting screw to cure occasional tripping.

Despite all its trials, the chronometer responded to my ministrations and ran sweetly on a full wind. After some adjustment, it ran over 20 days with a mean gaining error at room temperature of 0.9 seconds per day, with the mean of the deviations from this mean error being 1.03 seconds. Making a case for it took a little while, but it seems happy enough in what I was able to achieve (Figures 3 and 4).

Case 3 4ths view

Figure 3: Exterior of new case.

I was able to use some Hamilton handles, but the gimbals lock and the brass corners I had to make myself.

Case open from L

Figure 4: Chronometer in new home.








27: Fusee chain substitute

13 02 2017

A few weeks ago I received a chronometer which seems to have been heavily damaged by dropping, possibly while out of its bowl . The upper balance pivot was broken, the upper jewel of the third wheel was absent, or rather, it was present in many tiny pieces, and there were burrs raised on a few adjacent teeth of the great wheel. The stop bar of the stop work was stuck and the maintaining power did not work. The hands were absent and so was the fusee chain.

I have covered re-pivoting in posts 6 and 7 and making hole stones in post 26, though when there is no end stone to take into account, the process of making one is a rather simple turning operation. A little careful filing removed the burrs on the great wheel. Dismantling the stop work and maintaining power revealed dried-up grease, so that merely cleaning the parts of all old lubricant and replacing it with new brought these back to life. That left the hands to make and the chain to replace. I will cover making hands in a later post.

To a mariner, “cable” used to mean “A large rope by means of which the Ship is secured to the Anchor.” When it became possible in the nineteenth century to make large chains economically, they began to talk of “chain cable”. While the weight lines in clocks were of gut, this presumably was not strong enough to use as the link between spring barrel and fusee, so chain was used. It may well be that the technology of the eighteenth century was not equal to making wire lines that were both flexible and strong, and since chronometer makers were nothing if not conservative, the use of chain continued to the end of chronometer manufacture. These chains, which look rather like miniature cycle chains, were made by specialists and the chronometer repairer limited himself to repairing a broken link, usually by simply removing it.

Such books as there are that tell of chronometer repair are silent on what to do when the chain is completely absent. I found that I could buy a replacement chain for a trifling US$190, but as readers will perhaps by now have gathered, I am no spendthrift, and I began looking for an alternative to chain. The problem resolved itself into several requirements. The replacement has to be strong enough for its task; it has to be flexible enough; and it has to be possible to attach it to the barrel and fusee without special techniques. It also has to fit in the pre-existing groove of the fusee. These groves are always flat-bottomed so that conversion to a round cable is possible.

Less than a kilometre from my house is a famous game fishing club from which tormenters of large fish set sail. It is considered unsporting of the fish to bite through the trace of the hook and so it is often nowadays made of steel wire. A visit to a local supplier of fishing tackle provided me with a metre of stainless steel wire 0.8 mm in diameter for the sum of NZ$1.20 and I thought my problems were at least partly solved.

I carved hooks out of some scrap roofing iron using a piercing saw and files, and immediately encountered some problems: I could not solder the wire to the hooks, it was not flexible enough to allow me to pass it through a hole and bind it back on itself and when I made a little brass bush and squeezed it on to the end of the wire, it projected enough to foul either the top plate or the ratchet wheel of the maintaining power or the centre wheel arbor. By creative bodging I managed to overcome this, only to find that the wire was almost too stiff to be able to fit it. I then recalled that many years ago a kind person had given me a box of bits and pieces that he had bought at a deceased clock maker’s clearance auction with no clear idea what to do with them. Among the bits and pieces were several pieces of flexible wire and one length of the correct diameter to fit between the cheeks of the fusee groove. It appeared to be of multi-stranded steel, was very flexible; and it could be doubled back on itself with relative ease.

A test of its solderability revealed that it was plastic-coated and when I burned off the coating and scraped the wire clean, its fine structure could be revealed and contrasted with the relatively coarse structure of the stainless steel trace wire ( middle, Figure 1). The figure also shows how I at first anchored the trace wire to a hook with a bush.


Figure 1: Stainless steel trace (left and centre) and multi-stranded flexible steel (right)

Attaching the cable to the barrel with a normal-shaped hook was relatively easy, using techniques well known to a mariner setting up standing rigging. In my case, after doubling back the wire on itself I held the bight in a small vice while binding it with some fine nichrome wire that I had bought for another purpose. As Figure 2 shows, the free end  just clears the top plate.


Figure 2: Attachment of cable to the barrel hook.

Attaching the other end to the fusee hook required some modification to the normal shape of the hook, as otherwise the standing end of the wire would not lead into the beginning of the fusee groove. Thus, I made the hook deeper than usual so that the end of the hook could be twisted through 90 degrees and bring the cable in line with the fusee groove (Figure 3).


Figure 3: Attachment of cable to modified fusee hook.

I have not been able to find a supply of similar wire. Phosphor bronze wire 1.5 mm in diameter is available but is too big. I tried unravelling  a sample of the wire and using only three of the five strands, but it stubbornly refused to be pulled straight. As it was hard drawn, I tried annealing it in a gas flame while pulling on it, but it broke under a very moderate stress. Clock fusee chain is too wide and pocket watch chain is too short.so I am very glad to have been able to find just the right diameter.

After a thorough overhaul and repair, the chronometer sprang into life with a very strong balance motion. I have gradually regulated it  over three days and it is currently losing at the rate of 1.3 seconds a day. I hope that with a little more tweaking it will do a little better.

If any reader knows where to obtain flexible, multi-stranded steel or phosphor-bronze wire of between 0.8 and 1.0 mm in diameter, I would be glad to hear of it.



26: Making a hole stone.

24 01 2017

When chronometers get dropped sometimes it is not only the pivots that suffer. The jewels, often known as “stones,”  sometimes get chipped or cracked. Continuing to run a chronometer with a cracked stone is to invite the pivot to develop grooves and eventually the pivot gives way. A half-dead pivot can be seen at Figure 2 of post number 5 of April 2013. When a stone is chipped and the hole in it becomes irregular, often the chronometer will not run at all, or if it does, may have an irregular rate, or only run one way up. Unfortunately, replacement hole stones in chronometer size are hard to come by for the occasional repairer, though occasionally mounted stones turn up for Soviet chronometers on e-bay – at a price.

Early chronometers had few jewels, but ran well with the pivots running in the brass of the plates, or in bushes let into the plates. A few days ago, I became impatient waiting for a replacement to arrive and thought to make a replacement out of brass. Phosphor-bronze might have been better, but I didn’t have any. The main problem is in drilling the tiny hole required, 0.2mm diameter for the hole stones of the Soviet MX6 chronometer’s balance and escape wheel staffs. It is now possible to buy tungsten carbide drills of this size, but they were intended for drilling fibreglass printed circuit boards using specialised machinery. In the amateur or repairer’s workshop, where the tailstock of the lathe or the drill chuck may not be perfectly co-axial with the lathe axis, drilling a tiny hole off centre results in the drill wobbling and then breaking. In what follows, I have tried to show a method that works for me in producing emergency replacement hole stones, but first, a sketch of an MX6 holes stone. I have not dimensioned it fully, as other makes of chronometers may well need different thicknesses and diameters.


Figure 1: Sketch of MX6 chronometer hole stone.

The upper face of the stone has an annulus  excavated for a depth of about 0.3 mm around a central island and the face of the island is relieved by a “whisker”, perhaps 0.05 mm, so that when the end stone is in place, there is a slight gap between them to allow oil to run by capillary action and form a reservoir. The lower face is also relieved by drilling in stages so as to leave a hole 0.20 mm in diameter and of about the same length. The point of a larger drill leaves a conical lead-in to the hole, to help the end of the pivot to find its way and then an end mill gives a flat-bottomed hole surrounding. it.


Figure 2: Facing end of the blank.

A scrap piece of brass is first turned down to a close fit in the hole in the plate, in this case 5.00 mm diameter and then carefully faced (Figure 2)


Figure 3: Excavating annulus.

Then the annulus is turned away with a round-nosed tool (Figure 3). Note that the tool has negative rake, which gives a very fine finish on brass. It doesn’t matter how deep the excavation is, as long as it is much more than a whisker deep! Then the face of the island is relieved by 0.05 mm, as noted above and the blank parted off  a little over-thick (Figure 4).


Figure 4: Parting off.

One must now digress to make a shallow socket in the end of a scrap of brass bar (Figure 5).


Figure 5: Drilling a socket

The diameter of the scrap doesn’t matter as long as it is larger than that of the blank, but the size and the depth of the hole does matter. Unless a drill is perfectly sharpened it is likely to drill over-size, so I suggest making a start with a centre drill, followed in this case by a 4.7 mm drill. Then even a 5 mm drill that has slightly un-equal cutting lips will follow the hole so-formed and be of the correct size. The finish of the walls may not be great, but the size will be right. I made mine a mere 1 mm deep. Before proceeding further, measure the thickness of the blank so as to know how much to take off the thickness. As the faces of parted-off material may slightly concave or convex, measure near the edge, as in Figure 6.


Figure 6: Measuring thickness of the blank.

Then the socket is heated gently and annointed with flake shellac to leave an even coating inside. While the shellac is still liquid, the blank is firmly seated in the socket and the whole allowed to cool, when the socket and blank can be replaced in the lathe.


Figure 7: Facing underside of blank.

After careful facing and removing metal to the correct thickness (Figure 7), drilling can begin. Here, careful depth control is essential, using the graduated thimble on the tail stock quill rather than a ruler. A dial gauge could also be pressed into service. Carbide drills are usually sharpened by the four facet method so that they are self-centring, but to make sure, make a dimple with the much more rigid centre drill and then feed in the 1 mm drill carefully to a pre-calculated depth (Figure 8).


Figure 8: Drilling with 1 mm carbide drill.

In the example shown, this leaves only 0.2 mm for the very slender 0.20 mm drill to tackle and there should be no problem of it wandering off centre , since the larger drill leaves a conical lead in.(Figure 9).


Figure 9: Drilling with 0,20 mm carbide drill.

Although tungsten carbide is a very strong and dense material it is also brittle, so the lathe has to run at high speed and the feed has to be slow to avoid breakages. Once the hole has been drilled successfully, it remains only to use a 3 or 4 mm end mill to form the relieving step. The depth of this must also be calculated, as it is important to leave the conical run-in area to guide the pivot into the 0.20 mm hole. The new stone is now separated from the socket by heating and given a soak in alcohol to remove all traces of shellac, after which the thickness can be checked. It is a good idea to brush the floor around the bench when separating the stone, to make it easier to find it when you drop it on the floor.

Figure 10 is a composite photograph to show both sides.


Figure 10: The finished stone.

I don’t know how long this improvised lower hole stone for a chronometer escape wheel pivot will last, but at the moment, the instrument is running well and heading for a good rate. I have a French drum clock that was one of  my grandmother’s wedding present in which the relatively fine pivots run directly in the plates and there is no sign of wear at the low power end of the train after over a hundred years of continuous running, so I hope that my stones will also last over a hundred years. If they do not, they are easy to replace.