CNC Clock Build W.I.P.

[J-G]

18th Jan
Quite a productive day today - how much that is down to the extra light in the workshop is debatable! - I got the Main Drive Gears and Ratchets finish sanded/sealed/MC waxed and polished. It never ceases to amaze me how good it feels to handle nicely polished hardwoods, the Ratchets are made from African Leadwood and the finish feels like silk. Not that the end user will get to handle the 50mm Ø face now it is glued to the Winding Drum. Equally with the Hard Maple Main Drive Gear although there is enough surface area to handle, it’s not something that would be expected to be touched but there is a certain satisfaction in knowing that it could be.

Regrettably it’s difficult to show the polished finish in a photo, though there is just a hint of a highlight on the rim of the Ratchet in Fig-197.

Now that I have the Drive Spindle assembled I’ve proved that the Ratchet & Pawl system really does work.

I next turned my attention to the First Train Spindle (this carries the Hands) and was somewhat miffed to find that one of the four which I’d glued up yesterday had the 60T Gear set at about 3° out of square The only solution was to saw through the spacer, re-cut the recess on both the 60T & 32T gears then make a new spacer. Fortunately I didn’t have to re-make the Gears!

Once I’d re-made and glued this pair of gears - making very sure that they were running parallel - I had to leave them for the glue to cure so turned my attention back to the Main Drive Spindle. I said some time back that I wasn’t sure whether I should buy a commercial Winding Handle or make one from scratch.

Part of the reason for this indecision was the fact that they were available at about £8.50 but also that it is quite a complex item. So, I looked again at the commercial product to both check the cost but also to re-evaluate the design. I soon discovered that the price today is nearly £14 - no brainer! - so I set about designing my own.

The ‘norm’ is to machine a square on the end of the spindle (Fig-198) but that does cause a problem when you don’t have access to a square broach to cut the mating square in a ‘Key’. However, thinking outside-the-box, there is no reason that the drive has to be square, it could be two flats (Fig-199). Yes, a square would keep the key in place but a matching slot can easily be milled in a round bar.

One problem might be that there would be a tendency for the key to slip off the spindle. A solution would be to put a ‘sleeve’ over the slotted round bar. A further issue might be the ‘thin’ prongs left after the slot had been milled. I think I’ve solved all these concerns, along with some others that I didn’t foresee until I set about drawing both a SketchUp model (Fig-200) & a CorelDRAW! technical drawing (Fig-201). I wanted the ‘Knob’ to be captive but ‘loose’ so I had to make a small change from my original design which had the Spindle (116) as a simple 6mm Ø peg with an

M4 thread on the bottom which would also act as a clamp for the Arm (114) - to incorporate a hexagon collar for a spanner. On further consideration I may well Drill & Tap the Arm so that the thread on the Spindle goes through it.

Since I don’t have any 10mm A/F Hexagon Stainless Steel ‘on-the-shelf’ I’ll have to visit a local metals warehouse tomorrow.

To get around the ‘thin’ pegs I’ll make the slot 5mm wide rather than 6mm.

19th Jan
No real work done on the clock today, Stainless Steel Hexagon bar bought along with similar Brass which I also needed to make the Vertical Adjuster Screws, other than that, all I got done was the small modification to the Winding Handle drawings - ‘Life’ just got in the way

20th Jan
A long day in the workshop but I only got a prototype of the Winding Handle part made. I still have to finish the base of the ‘Knob’, fully insert the ‘pronged’ part and pin it in place and make the brass plug for the centre of the knob.

Making the prototype has brought some issues to the fore and the next three will be marginally modified. Specifically, the knob has a brass sleeve which is a straight through tube but I now see that it would be better to make it blind so that the fixing does not rely upon a thin sliver of Walnut to retain the Knob on the Spindle. This becomes possible because the Sleeve can be affixed to the Walnut Knob with CA glue.

21st to 23rd Jan
The Winding Handles have all been completed, again with small - practical -modifications.
The original design had a ‘pin’ holding the slotted insert in place but since I don't have any small taper pins I changed that to an M3 grub screw. I also decided to drill & tap both ends of the Arm rather than rely on a simple ‘clamping’ against a milled flat.

The Brass i

ncerts proved to be simple and they are glued in place making the Knob ‘captive’ on its spindle so it is free to rotate but I had to make a thin 10mm A/F spanner to tighten the Knob Spindle to the Brass base.

With 10 components, the Winding Handle is probably the most complex part of the whole clock so no wonder it took be the best part of three days to make the four. Far too much of the time was spent in tapping the Arm – Stainless Steel is much more difficult to work than EN1a and fortunately I didn’t break any taps! I did break 3 Carbide drills and even a 4mm end mill though.

24th Jan
With the completion of three more brass Vertical Adjustment Screws today (now I have the Brass hexagon bar) I think all the components, other than the Weight, are complete so I can’t put off applying the finish to the main frames any longer. The faces had already been sanded to 320g but I hadn’t done anything to the edges so running those against a flap-wheel, some hand sanding where I couldn’t get the flap-wheel in, before a coat of Sanding Sealer followed by a de-nib with 1000g and a final MC Wax, has made them ready for gluing up with the Dial stand-offs and Spacer Brace.

 

[donwatson]

Thanks JG

 

[J-G]

25th Jan
Today was spent going through the various components finish sanding, sealing, waxing, polishing and finally allocating to a particular build. This means that all the less than perfect - or even outright ‘damaged’ but still functional - parts will all be on the one clock. The sorting has brought to light the fact that I’ve ‘lost’ a few components that I know have been made but I’ve been careless and dropped them whilst moving them around the workshop. I now have four containers to store the smaller parts so hopefully won’t lose any more!

26th - 27th Jan
Working through ‘fiddly’ bits and pieces which of course take far longer than it seems they should. The Pendulum Rods are now ‘pinned’ to the Escape Wheel Latch/Lift and screwed to the ‘Bob’.

I’ve now started real assembly - using Glue - ie. no easy return to the previous state.

This has been a simple matter of fitting the various pads to the frame though I have also built the Escape Latch support and fitted the Frame Spacers so I can at least show an assembled frame complete with Pendulum. The Bearings are all fitted so hopefully I’ll get the spindles & gears fitted tomorrow.

29th - Jan
Nothing done yesterday but today I started by gluing in the Frame Brace and Spacers to the Rear Frame using the front Frame with the Spacing Rods as a clamping mechanism.
Once the glue had cured I could take the frame apart again and realized that I needed to make sure that the Latch & Finger Support was correctly aligned. This is the curved part at the top on which the Pendulum ‘Finger’ and the Escape ‘Latch’ are attached and there is a means of adjusting its position to bring the clock ‘In Beat’ but naturally it would best to have it in a central position to start with. This is achieved by using the two M3 grub screws either side of the Rear Frame which act upon a Brass plate - - - a drawing paints a better picture! - Fig-206.

To achieve this ‘balance’ I had to make sure that the Frame was held exactly perpendicular to the axis of measurement so I lashed up a jig (Fig-207) with which I could measure the position of each ‘pivot’ accurately. That is between the two Red lines, the distance along the black lines must be equal.

That is, under ideal conditions,

the beat will be Tic-Toc-Tic-Toc not Tic—Toc-Tic—Toc-Tic—Toc - - - - - - or any other variant!

The problem is that once the Clock is in use I have no control over the particular location and a fine adjustment becomes necessary. This is also why there is a fine adjustment for the Pendulum length since Gravity is not an absolute constant geographically.

 

[donwatson]

Thanks JG

 

[J-G]

The first ‘Build’ has revealed some issues - as was to be expected - nothing particularly onerous but certainly needing attention. There seems to be no problem with the gear teeth ‘binding’ but there is an issue with ‘wobble’ - that is with the Gears not being absolutely perpendicular to their spindles, but it may be more to do with warping or cupping of the timber.

This first showed itself with the Escape Wheel where I designed in quite a small space between the Frame and the back of the Wheel - 2mm - so as it is 120mm Ø and only 5mm thick, a small warp of 1mm should not be unexpected but couple that with the fact that the head of the Latch & Finger Support Clamping Screw stands about 1mm proud of the frame and there will be no clearance

An easy solution to this was to re-cut the Support Pivot Bush recess 1mm deeper thus putting the screw head at least flush with the frame. In Fig-208, which is a photo’ taken from the top of the frame, you can see that I haven’t given myself a great deal of lea-way.

With hind-sight, it might be better not to rely upon the gears remaining ‘flat’ and finish machine them once they have been assemble on their spindles. I will have to dis-assemble the clock at least once because I still need to cross-drill the 1st Train Spindle for the pin to drive the Hour Hand Gear and the position of that needs to be determined empirically, so I still have the option to true-up the Escape Wheel - and of course there are three more clocks so I can take this new knowledge into account before I do the next build.

After the minor adjustments I was pleased to be able to turn the Main Drive Wheel and see all the Gears & Escape Wheel turn freely.

What is less gratifying is that I can only do so by pulling quite hard on the Main Drive Gear rather than pulling on the Cord. This may mean that the weight needed to drive it may be much higher than the estimated 2kg

30th Jan
I got my Grandson to pull on the cord with me holding the Clock Frame and it seemed easier to get the Gears moving and I’ve now found a space in my workshop where I can fix the Wall Plate and therefore ‘hang’ the Clock for the next series of tests.

My first test was with a 1.7kg bottle of water which needed a little persuasion (by ‘flicking’ the Escape Wheel) to get any movement but it was no more than about a ¼ turn of the Main Drive Wheel. Changing to a 3kg bottle did get it to completely unwind the winding drum but it took between 4 and 5 minutes to do so. I also became aware that with that weight there is a tendency for the right hand wall anchor to lift out of the key-hole slot in the Wall Plate so I’ll need to design a locking mechanism to stop that.

Finding a lump of steel weighing about 2kg I lashed that to a (currently unused) part of my Myford 7 which brought the weight up to just over 4kg. This was less than successful since the left hand Wall anchor broke ! The main reason for this was that I had been less than fastidious when gluing it to the frame so it was only held in by a tenuous 5% of the circumference with no glue on the shoulder. The fact that I didn’t remove the wax polish first may also have been a contributing factor!!

Due to the right hand Wall Plate button lifting, I needed to design a locking device. It made sense to devise a ‘cam’ locking which could be lifted out of the way to allow the clock to be mounted and locked down again once in position. As the Brass ‘Key-hole’ plates are held in a recess in the Wall Plate with two screws it also seemed convenient to use the top one as the Cam axis by fitting an M4 Thread Insert where it is currently a wood-screw. Fig-209 shows the right hand end of the Wall Plate with an outline of the Cam, open & locked. This was a simple matter to CNC out of 3mm thick Tufnol but I still had to do the job twice through stupidity! ie. not checking the location of the blank after re-clamping. Now I have to re-make the Mounting Block which broke so that I can test how effective the Locking Cam is.

31st Jan
Since the new ‘Locking Cam’ is only 3mm thick and the recess for

the screw head can therefore sensibly be only 1.6mm deep, I can’t use standard button head screws, the standard head is 2.3mm thick and 7.7mm Ø so I had to modify them – a simple matter on the lathe but still taking time.

Cutting the hole for the M4 Thread Insert was also a simple matter, using the existing countersunk screw-hole in the Brass plate gave ample positional guidance and once that had been opened up to 4mm I could remove the plate and drill the 5.5mm Ø hole and 8mm counterbore needed for the Insert. One small bone of contention is the depth of the hole - - - I need the screw to ‘bottom out’ before the Cam is clamped tight and can’t swivel and that could be a matter of just a few tenths of a millimetre. There seems to be two ways to deal with this, either trimming the length of the screw or adding a small piece of steel or brass to the bottom of the hole.

Before I deal with that, I’m thinking about why it seems that I need a 4kg weight to drive the clock. One thing that occurs to me is that there may well be a pressure on the ends of the spindles created by the spacers being just a little short - or possibly that the Bearing recesses have been cut a little too shallow - either scenario would have the same effect. The fact that the Frame Clamping Rods seem just a bit too long seems to bear this out.

An easy way to test this would be to add a ‘washer’ under each Spacer. It cannot be too thick so a sample of Formica at 0.7mm thick might just be enough, but I could add two if one made little improvement.

An interesting morning, discovering some of the errors made early on in this project! The addition of the thin washer did wonders for the ‘weight’ issue – along with a small adjustment to the Escape Wheel to correct some of the ‘wobble’ - and I had a complete unwind of the drum with a 1.5kg weight.

I said ‘some’ of the wobble - - the Pins were meant to go in 4 of the 5mm thickness but I’d obviously been a bit too heavy handed and taking just about ½mm off exposed some of them. In case that ½mm wasn’t enough, I decided to take a further ½mm out of the Spindle shoulder. The potential repercussions of this would be a possibility that the Pinion wouldn’t align with the Gear but since they are both 5mm wide, pushing it ½mm along the spindle is not a serious issue.

After re-building the clock I mounted it on the Wall Plate and a further ‘free-fall’ test proved positive with no ‘rubbing’ between the Escape Wheel & the Latch/Finger Support.

Now came the acid test - with the pendulum.

Not 100% - in fact I had to increase the weight to 2kg before I got any useful power and even then it was very ‘hit & miss’ getting no more than about 15 seconds. After some investigation it seemed that there wasn’t enough power being past to the ‘Finger’ which is what the Pendulum needs. There was also an issue with the Latch not lifting quite enough to release the Escape Wheel. This last point is exacerbated by the fact that the Escape Wheel is not perfectly concentric with the Spindle - - it’s no more than ¼mm but that is a huge amount when the total lift on the Latch is only 1mm!

Stripped the clock again and mounted the Escape Wheel - on its Spindle - in the lathe and trimmed the top until all the teeth had been affected. I took off about 0.3mm from the highest tooth. Whilst I had it in the lathe I took the opportunity to apply some pressure to the Pins since some of them are just off being perpendicular.

With this ‘hind-sight’ knowledge, I think that rather than drilling the holes for the pins at the same time as I cut the teeth on the Denford, it would be better to leave that until the Wheel is mounted on its spindle and drill them on the lathe using an index wheel or on the Mill using a dividing head - it would be more work but guaranteed to be accurate. If the Escape Wheel were made of Brass I suspect that there wouldn’t be any problem.

Now I’ve done the same adjustments to all four Escape Wheels, it’s interesting to note that two of them had virtually no ‘wobble’ but one of them had nearly 3mm ! and that one is the first one I made! It was built from the 5 segments which I did thinking that it would be a way to eliminate some potential problems - obviously not!

I did lots of ‘fiddling’ with the Latch-Lift/Finger by adding bits to both and trimming the added bits back but still no real improvement. The one thing that did make a difference - and effectively changes the position of both the Finger & Latch-Lift - was to adjust the Pivot Screws. By screwing

them further in, both the Latch-Lift and the Finger are raised changing the relative position of the Finger ‘tip’ to the Escape Wheel Pins and this certainly had a positive affect and I got a Pendulum swing lasting about 4 minutes.

2nd Feb
After a few attempts at cajoling the Pendulum to swing longer by increasing the Weight to 3kg and further adjustments to the Pivot screws I took a more pragmatic approach and made a pencil mark on the Gear Teeth when the Pendulum stopped. By now I had fitted the hands so that I could judge how long the Clock had been ‘going’ and it had achieved over 7 minutes.

It was good to find that it stopped exactly at the same position on three occasions which meant that there is certainly some binding between the Second Train Gear & Escape Train Pinion so I stripped the clock yet again and trimmed the 10 teeth either side of the one that was marked. I also mounted the Gear & Spindle in the lathe to check that the O/D was concentric and it turned out that the tooth with the pencil mark was about 0.3mm higher than the lowest tooth and was soon trimmed to concentricity. Likewise, I tested, and corrected, the First Train Gear which had a smaller difference.

After re-building the clock and mounting it for further tests, adding the weight seemed perfectly normal but after a few swings of the Pendulum it plummeted to the ground!!! It was disappointing to discover that two of the retaining pegs on the Pawls had broken away allowing the Winding Drum to ‘free-wheel’.

Fig-212 is the outline drawing of the Pawl and Fig-213 is the broken Pawl.

Since the Pawl is such a small component I suspect that I could have foreseen this possibility - - it doesn’t make it any more palatable - - but I think that making the replacements from Tufnol may be a better option, though increasing the size of the retaining peg (al la Fig-212 dotted red line) could be a way to maintain the ‘Wooden’ element. I’ll make that decision tomorrow.

 

[donwatson]

As usual a big thanks for this JG. Working on this project looks as if you have hit quite a few running problems/glitches and I find it interesting to see how you solve them. Good ideas you have on problem solving, keep up the good work.
take care and stay safe
Don W

 

[Yojevol]

Your latest niggles sound very familiar JG. 90% of my similar problems were associated with the escape wheel where the available torque is at a minimum. I took particular note of your gearwheel having a 0.3mm ovality. I have experienced exactly the same error on my first attempt at gear cutting on my Denford. I ran a few tests to find that I am getting constant errors especially in the X-axis. I've now found that I've got a 0.3mm backlash in the leadscrew and measurable wear in the guides which is giving smaller errors in the Y direction. I've opened up a thread on the Denford forum which you may be interested in. Also seeking advice here. I'm making some headway in resolving the issues.
Brian

 

[J-G]

Working on this project looks as if you have hit quite a few running problems/glitches and I find it interesting to see how you solve them. Good ideas you have on problem solving

Thanks Don. When I started this WIP I did say that it would be 'my take' on what a WIP should try to do - that is, catalogue the errors made and show how they are circumvented.

I have no illusion that I'm immune from error and probably make more than most since, even though I do always work though potential issues by creating 3D drawing well before I cut any material, it is often not until the manufacturing stage that an unforeseen 'issue' comes to light.

In the case of a clock there are some components where the tolerance to error is extremely small and the Escape Train is certainly one of them.

The fact that the Pawls have shown that they need a much greater strength than I originally thought is just part of a learning curve and could only have been predicted if I had real knowledge of what forces would be likely to be applied, along with details of the shear-strength of potential materials, rather than a gut feeling that "Beech ought to be OK"

 

[J-G]

Your latest niggles sound very familiar JG. 90% of my similar problems were associated with the escape wheel where the available torque is at a minimum.

I wish I had the maths knowledge to be able to determine the amount of Torque available! I know it is only a matter of determining the Mechanical Advantage at each transfer of power and for each Gear/Pinion combination that isn't too onerous but when you get to the Latch/Finger/Pendulum arrangement that's a whole different ball game. Couple that with the fact that when I did an ONC as part of my apprenticeship, I only just scraped through on the 'Mechanics' section with a 39% being re-assesed to 40% - (the 'pass' mark).

I took particular note of your gearwheel having a 0.3mm ovality. I have experienced exactly the same error on my first attempt at gear cutting on my Denford. I ran a few tests to find that I am getting constant errors especially in the X-axis. I've now found that I've got a 0.3mm backlash in the leadscrew and measurable wear in the guides which is giving smaller errors in the Y direction. I've opened up a thread on the Denford forum which you may be interested in. Also seeking advice here. I'm making some headway in resolving the issues.
Brian

I had noticed your posts about the potential use of Nylon Rod as a replacement guide material but didn't have a reasoned opinion to contribute. I would say though that I would be cautious about 'gluing' any material into the slide carriers. A mechanical fixing can always generally be reversed in the event that it is ineffective.

In my case I don't think it - necessarily - is the same, and it's not 'ovality', which would mean that the error would have two high points at 180° apart and similarly two low points. Mine is more a fact that the centre hole is not 'centred' and since the hole was machined on the lathe with the Gear being held in a pre-machined, and therefore total concentric, set of jaws, I suspect that it is more to do with the natural movement of the Maple.

The link to the Denford Forum is useful thanks - I hadn't looked for one - and I've already down-loaded the Feeds & Speeds .pdf which I'll look at shortly.

 

[J-G]

3rd Feb
Making small adjustments to existing drawings within CorelDRAW! is very easy but still seems to take forever. Adding the bulge indicated by the dotted red line in Fig-212 appeared to be the best option - whatever the material I choose - and by adding that it would also be necessary to increase the bulge at the top so that the weight is sensibly distributed because the Pawls are only kept in contact with the

Ratchet due to gravity. In Fig-214 you can see the outline of the new Pawl in red compared to the old Pawl (dotted blue). The geometry of the main part is naturally not changed.

Considering the options open to me as far as the material with which to make the new Pawls, I was concerned that - now I know the Drive Weight is likely to be in excess of 3kg, (I had originally anticipated about 2kg or less) - Mapl

e, Beech, Ash, Oak, Elm etc. are unlikely to be strong enough to withstand the forces involved for very long. Other options are, Tufnol, (a Phenolic Resin Laminated material which essentially has no ‘grain’), Brass or Aluminium.
The later two probably could be cut on the Denford but I’ve not yet dipped my toe in that water so I settled for Tufnol and the result can be seen in Fig-215 where the Pawls are mounted on the Main Drive Gear.

As there had been damage to the Pawls I was not at all surprised to find that when I dismantled the Winding Drum from the Main Drive Gear, there was also damage to the Ratchet. That is quite difficult to photograph so I’ve done some enhancements to Fig-216 where the Yellow line shows where the Ratchet teeth should be (were!) and the red lines indicate the damage which amounts to three of the teeth.

The Ratchet is already glued to the Winding Drum so affecting a repair to this is going to be fraught with issues.

I could make a new Ratchet and destroy the existing one by holding the Winding Drum in the lathe and machining it away but I think that inserting Brass ‘Teeth’ would be preferable. These might be glued into radial s

lots which could be cut using an engraving MOP using the blue lines in Fig-217.
One problem with this is that I need to find a way to hold the assembly which is the Ratchet & Winding Drum on the Denford Table so that the Ratchet is uppermost but the bottom of that assembly is only 12mm Ø.
The centre of the Denford Table has an M6 threaded insert so I made a locating peg to fit the centre of the Winding Drum with an M6 thread on the bottom and an M5 thread at the top. This would ‘locate’ the Ratchet but wouldn’t hold it well enough to handle cutting the slots for the Brass teeth so the possibility for it to move would be high.
Next I bored a close-fitting hole in a block of Oak which can be clamped to the table and milled a slot into which the M3 Grub Screw which clamps the Winding Drum to the Drive Spindle can fit thus arresting any potential movement.

4th Feb
The Ratchet is now positioned on the Denford Table and Fig-218 shows it with a 2mm Ø burr at X0-Y0 but I’ve just realized that I haven’t taken account of the precise orientation of the Ratchet Teeth . . . . . This is going to be important because initially I’m thinking of just inserting Brass teeth at the position of the three broken ones might do the job.

. . . . . . Now I have correctly aligned the block, I’ve noticed that a fourth tooth has some damage so that will also be replaced.

One of the issues I’ve become acutely aware of, as far as CNC Machining is concerned, is having the work-piece held firmly. Any possibility that it could move during a cut is pretty much certain that it will move and cause damage to either the work or the cutter - or even both! So, I was well pleased to find that this set-up worked very well. In Fig-218 the clamps are gripping the Oak block by about 2mm - they were firm enough but when I re-aligned it to correctly orient the Ratchet Teeth I increased this to nearer 6mm and it was solid as a rock.

Using a DOC = 0.5mm limited the stresses and the four slots were perfectly positioned. Making the Brass teeth was just a ‘fiddle’ at 5mm x 9mm x 2mm and one end rounded over but they were easily

glued into the slots with CA, left for 10 minutes before mounting the Drum in the lathe and taking cuts of about 0.2mm soon got all the Brass Teeth at the same height as the wooden ones.

This of course left them 2mm wide with no relief and filing the back away when the length of ‘stroke’ for the file was only 7mm was interesting.

Doing ‘repair’ work might be interesting in-as-much-as it is always a one-off job and often needs ingenuity, but it does take an inordinately long time! It may well be that I could have re-made the complete Ratchet in less time but the chance of similar damage would still be present.

 

[Yojevol]

JG I'm surprised you think the required weight is going be 4Kg or more. That is going to add directly to the loads on those pawls and ratchet teeth. Have you tried assembling the clock in stages and test running it at each stage? This gives a better idea of where any unwanted friction is might be arising. At each stage of assembly the gear train should be stopped and started to simulate normal running speed. Slow intermittent motion is much more onerous than just letting it freefall.
Brian

 

[J-G]

Thanks for the input Brian. With my last run, most of the tests were with 2.6kg and was getting 7 to 8 minutes, only later moving to 3kg, and suspect that the damage was caused earlier so after the latest round of adjustments I hoped to get more.

As you'll see from the latest post, I've not got as far as re-building so no further real tests have yet been done.

I can't quite see what you mean by 'running at each stage'. Without weight, I have both 'spun' the Escape Wheel and pulled on the Main Drive Gear - to also make the Escape Wheel spin (both equivalent to 'free fall') but I've also gently rotated the Second Train Gear which I suspect is more akin to the 'normal' action and didn't 'feel' any resistance at a specific point.

I have tested each Gear/Pinion pair but not with the frame firmly bolted together so that may well be a sensible ploy.

I've also added weight to the Pendulum Bob but there didn't seem to be any benefit to that.

 

[J-G]

5th Feb
As suggested by Brian, I made a point of testing the ‘free-flow’ of each pair of Gears/Pinions by applying as little pressure on the gear to just make them move and therefore to see if there were any further ‘high spots’ which would be likely to need more torque to overcome the resistance. As I found them, I marked the teeth with a pencil and judiciously trimmed the flanks until such time as I had a full rotation with no binding.

Spent most of the day either making very small adjustments to either the Pendulum Pivot height, yet more judicious trimming of the tooth flanks or simply pondering over what else might be inhibiting more that a couple of minutes continuous run.

Finally gave up and got on with ‘life’

6th Feb
First thing I did today was check the Pendulum Pivot height and a very small adjustment seemed to show that the clock would run for over 20 minutes (using a 2kg weight) IF I constantly made sure that the Latch remained in position - that is centred over the Latch Lift. There was a tendency for it to move backwards and eventually get out of the reach of the Lift.

The Latch is constrained on a 2mm pin with a friction fit Acetal ring acting as a retainer but it obviously has to be totally free to move on its 2mm Ø Pivot Pin so the hole is 2.1mm. The Pin is held in the support by friction since the hole was drilled at 1.9mm but if the hole is even marginally out of square then there will be a natural equilibrium pushing the tip of the Latch out of position. This ‘tendency’ can be eliminated by a small re-alignement of the pin which is what I did with a pair of Pliers!

Of course this meant that I had to dismantle the clock again so whilst it was dis-assembled I took the opportunity to make other adjustments. I particular I took a further 1mm from the shoulder of the Escape Wheel Spindle giving yet more clearance between it and the Latch Support and better aligning the Latch with the centre-line of the Escape Wheel. I also sanded the flanks of the teeth on the Second Train Gear both sides which - since I’m using 1000g abrasive - is effectively polishing them but doing the same to the mating Pinion is very difficult because that is glued in place such that there is only about 10mm space so I can’t do it with the 50mm Ø disc. - - - - I think I’ll have to make a small round abrasive ‘stick’ that will fit the Dremel.

Before that, I re-built the Clock, hoping that I would see at least 20-30 minutes running - - - - No more than 2 - - - - even after re-tweeking the Pendulum Pivot height.

It seems that the 2kg weight is insufficient but when the bottle with 3kg fell to the floor, it ruptured and I don’t have another bottle large enough.

When it stops, the Pendulum is still swinging for some time but the Escape Wheel is stationary so that indicates that there is still some binding between the Gears/Pinions so ‘polishing the Pinions’ could well prove to be the answer. It may also be that the Main Drive Gear and the First Train Gear that is driven by it needs ‘polishing’. . . . . . .
After yet another strip-down, ‘fettle’ & re-build I still can’t get a regular long run I did have one session approaching 30 minutes but most trials ran between 1 & 8 minutes but the biggest disappointment is that there is no consistency. I’ve done tests with and without the Pendulum ‘Bob’ - just using the brass adjusting screw to provide some weight. This effectively shortens the Pendulum length to well below the 1 metre but I wanted to know if the weight of the ‘Bob’ was too great (it may not be), and I’ve also added a lump of steel weighing 1kg to the Weight making it 3kg overall.

I was able to take a Video lasting nearly 3 minutes (it can’t be added to the .PDF of course) but I had to install a Video Editing program to turn it through 90°. Due to its size I then had to trim it so there is just one minute.

 

[donwatson]

Quite a bit of work there JG. Much the same place I was when I abandoned my efforts to make a clock. There seemed to be no consistency in what happened and when it happened. Wish you much good luck in sorting it out.

 

[J-G]

17th Feb
Though I haven’t reported my activity for a fortnight, I’ve not been idle - though I have spent more time thinking than doing!

Yesterday I had the clock going for nearly 12 hours and it only stopped because the temporary weight reached the floor. Today - up to now - it has done nearly 10 hours so I’m satisfied that the changes I’ve made latterly have been useful.

The main change has been to the Winding Drum. My original design called for 25mm Ø but I’ve now made a secondary Drum 50mm Ø. My calculations show that this ought to provide twice the amount of torque to the Escapement Finger but will of course halve the time between winds - at least I have a working clock!

I also re-cut the 64T 2nd Train Gear - somewhat of a challenge since I had to very accurately re-mount it (them) on the Denford (sorry didn’t take photo’s) making sure that the teeth were correctly oriented. This entailed making a jig to screw into one of the locating screw inserts in the Denford table and four clamps to make sure there was no movement. I only took off 0.2mm from each tooth flank but I also cut the root 0.5mm deeper.

The Red Line in Fig-220 is the new profile and the Blue Line is the original. I suspect that removing the

‘bulge’ in the addendum has had the most impact.

With the new profile, and the original 25mmØ Winding Drum, I was still getting less than 5 minutes run time with seemingly random stopping positions.

Now I have to turn my attention to the real ‘Drive Weight’.

I had plucked ‘about 2kg’ out of the air as a starting point and with that in mind had determined to use steel inside a clear Acrylic tube. Now I’ve found that it really needs nearer 3kg - and I suspect that 5 or even 6 would have been necessary had I stuck to the smaller Winding Drum - I’ve been looking today at Lead Shot. As you may well understand, there are certain restrictions on the use of shot which is used mainly to fill shot-gun cartridges - or, in small amounts, for fishing weights. After some hours searching for ‘sensible’ sizes and prices it seemed unlikely that I could get the 12kg in good size shot - - - I have learned a great deal about the way that shot is measured though - - - one site even suggested that to use it for anything other than ‘Fly’ tying would be illegal ??

Eventually I came across a ‘Sash Window Weight’ available in 50mmØ, 600mm long with a 12mm bore and weighing 12.7kg at a little under £60 plus carriage (not determined). Further searching with this new possibility found one supplier at £51 delivered and another at £37 + carriage which I’ve asked them for.

This will be ideal since I won’t have to drill an 8mmØ hole through 150mm of steel and one weight will give me enough for all four clocks. I’ll still use the Acrylic Tube with Walnut caps to cover it and may even use ‘Plasticote’ to further encapsulate it, totally sealing it from any possibility of both tarnish and environmental hazzard.

18th Feb
This morning I could determine how long the clock had gone for because naturally the weight had dropped to the floor and it was 21h 25m . This is with a drop of 1.3m whereas the design height should provide 1.5m, in which case I anticipate that I will get close to 25 hours. I can’t yet fathom how this can be since my original calculations predicted 26 hours but I thought that doubling the diameter of the Winding Drum would halve the run time - - I’ll have to do some more calculations!

Yesterday I was using a Weight of 3.048kg made up of a water bottle + a lump of steel and had also added an extra weight to the Pendulum Bob but in an attempt to remove some of the ‘trial adjustments’ today I took both of the extra steel lumps off and it looks as though it will work with a 2.2kg weight. I re-started it at 10am, it's still going and I've just checked and it shows 13:15 which is 5 minutes slow. I know that the Pendulum is too long because I have used a full 1m length of carbon fibre tube for the Rod but at least I now do have a 'benchmark' for proper adjustment. It's a pity that I can only test modifications in 'Real Time'.

The Lead Weight has been ordered and will be with me tomorrow so I can get on with making the other parts.

 

[donwatson]

Thanks JG, I thought you must have been busy making some adjustments.

 

[J-G]

I've just done an [Edit] to my last post and changed the .PDF so you may want to download the new version.

Good to be moving forward again

 

[Yojevol]

I re-started it at 10am, it's still going and I've just checked and it shows 13:15 which is 5 minutes slow

@J-G It's fairly easy to calculate from the pendulum formula the amount to shorten it by. If it's 5mins slow after 195mins running it will need shortening by about 26mm (assuming G in your neck of the woods is 9.81)

Interesting source of lead - ready made into suitable size cylinders. I've been using scrap lead flashing and casting it but finding something to cast it into is not easy. 50mm drinks cans are common but that was to large for my recent clock. A source of lead shot in small quantities is the aqua-lung diving community. You can get 2Kg packs which they use.
Brian

 

[J-G]

It's fairly easy to calculate from the pendulum formula the amount to shorten it by. If it's 5mins slow after 195mins running it will need shortening by about 26mm (assuming G in your neck of the woods is 9.81)

I'm sure it is Brian, I just haven't done the work

I know g at 52.5725ºN and 92m above Sea Level is 9.81269 so I have done that much previously.

Interesting source of lead - ready made into suitable size cylinders. I've been using scrap lead flashing and casting it but finding something to cast it into is not easy. 50mm drinks cans are common but that was to large for my recent clock. A source of lead shot in small quantities is the aqua-lung diving community. You can get 2Kg packs which they use.
Brian

I did look at a number of sources and after some research thought that it would be good to get AAA shot which is 5.16mm Ø but I could only find that in small quantities and always 'split' - presumably for quickly attaching to fishing line - the SCUBA community seem to use much smaller shot (#6 2.5mmØ) which has had more work done to it so is therefore more expensive (have I mentioned that I'm a tight-wad?).

The Sash Window community appears to be a better option with most offering 25 & 50mm Ø x 600mm long but also square section and I did find one company offering 'make-weights' which are shorter lengths which may suit you better - since I'm making four I can use most of the 600mm length.

Mine is coming from Mighton (the cheapest) - here's the link to their [Round] list

https://www.mightonproducts.com/products/sash-window-hardware/sash-weights/round-lead-weights
Having a solid lump with a pre-cast hole is a much better option for me and I'll make a sleeve using some Sycamore veneer that I have on the shelf so it won't matter if it tarnishes - it simply won't be seen.

 

[Yojevol]

@J-G The other method I have played with is to cut a length of flashing and roll it up tightly into a cylinder. Easier done with new rather that scrap sheets.
Brian