CNC Clock Build W.I.P.

[donwatson]

Thanks again JG. You are moving along very well with this project and I really enjoy what you are doing/explaining. Thanks again.
Don W

 

[J-G]

Thanks for that feedback Don. I sometimes wonder if I'm giving too much detail by treating the WIP too much like a diary but your comments seem to say that the detail IS useful. My feeling is that it might provide some ideas for other projects.

Your pic_3 does show a discrepancy between the size of the shaft and the hole and I agree that you're on a hiding to nothing with that set-up. Wood will move on a daily - even hourly - basis and the chance of the timber in the frame matching the movement in the spindle is remote at best so you have to allow a degree of 'slop' which is definitely what you don't need for the Escape wheel.

If you don't have the means to make steel spindles then the minimum I would consider would be Deep Groove Ball Bearings which you might be able to retro-fit to the frame.

 

[AES]

Well J-G, I can only speak for myself of course, but when you say:

"Thanks for that feedback Don. I sometimes wonder if I'm giving too much detail by treating the WIP too much like a diary ..... " I would say "Not at all Sir. Go for it, the more detail the merrier!"

As said, that's me, and is also my own "style", which I've found some members seem to like, others not. As ever in life.

But I will say that I do appreciate just how much time and effort it takes to get such posts ready, and at least from me "Thanks, please continue".

 

[J-G]

Thanks AES - that is useful feedback as well - and though it does take time, I don't consider that wasted, sometimes I'm ahead of myself and just jotting down what I 'intend' to do will help to determine what I actually do.

A bit like learning anything (in my case often music/lyrics) by physically writing the lyric down it has to go through a number of senses and being a Tenor who doesn't read the Bass Clef well, the mere fact of resetting it to the Treble Clef aids the learning.

 

[AES]

Funnily enough J-G, I (try to) sing Bass-Baritone in the choir I belong to. I'm not clever enough to do all that music transposing stuff, but I DO write out the lyrics - often several times.

Also in our choir we often sing foreign language songs, and especially for French lyrics, (at which I'm absolutely useless), I make my own "phonetic English" version of the French lyrics. You should see the look on my (Swiss) choir colleagues' faces when they ask to look at my own (English phonetic) "special cheat sheets"!

Thread drift (of course) but so far all my WIPs have been done during/after the actual work. "Pre-writing" at least some ideas may well make some sense for me, thanks.

 

[J-G]

Thread drift is de rigueur! I'm in three choirs - though not rehearsing at the moment due to Covid restrictions - and have a go at Italian, Latin, German, Russian, Welsh - even Dutch !! - but don't give me French - there are a few pieces (Cantique de Jean Racine...) but in general

One choir has a concert planned for July next year in Dresden as a repeat of one where Dresden's Neuer Chor joined us in Coventry Cathedral in 2019.

Glad you find the concept of "Pre-writing" of interest, Today Yesterday I started re-looking at how I'm going to make the gear blanks and the write-up for that is pretty much all 'before the fact'.

 

[donwatson]

Great comments chaps, I think a bit of forward thinking does no harm, and indeed will aid the the process to run a bit more smoothly at the end up.

 

[J-G]

I spent all day yesterday -- and a lot of Monday -- working out the segment detail for the Escape Wheel. During the Clock 'General Design' process I was only concerned with the overall size and shape of the Gears so didn't go into the nitty-gritty detail of how the blanks would be created. I knew that they would be segmented since I wasn't using plywood and at only 5mm thick solid hardwood will not remain flat enough at 120mm wide.
I alluded to this in WIP-7 but hadn't taken on board how it would be achieved. WIP-13.PDF attached.
=======================================================================
12th Oct
I’ve been putting off the preparation of the Gear Blanks for too long. Waiting for inspiration to some extent and trying to persuade myself that trying to minimize waste could be a false economy (and I can’t get to select the Walnut until Friday). I’ve already cut two pieces of Maple out of a 15mm thick billet that are really a bit too close to 5mm thick (the finished size) and all I have left is a 2-3mm sliver. The next strip I cut will be down the middle so there will be some ‘meat’ to provide an easy finish trim.

The Maple I have is 48mm wide and was bought with the intention of making up a 5 segment ‘ring’ plus a centre ‘boss’ - see Fig-37 in WIP-7 - but I now see that 48mm is insufficient to cover the 72° necessary to have the grain running in the correct direction ( it would need to be 83mm wide) so today I did some more work in CorelDRAW! & SketchUp looking at a 10 segment option. This is for the Escape Wheel which is 120mm Ø - ie. the ‘worst case’ situation - the smaller gears won’t be a problem. I also have Maple at 64mm wide but only 10mm thick which is OK for the 72 T gear which is 8mm thick (but will still need 6 or 8 segments).

Fig-66 is a screen-grab of the SketchUp image in which I’m testing how a 10 segment option would work and Fig-67 is the next iteration done in CorelDRAW! From that I can move the segments about to eventually create the .DXF file which will be used in CamBam to create the G-Code.

. . . . . . . .

. . . . . . . . . . . . Fig-66

13th Oct
I’ve looked again at this problem, so Fig-68 is the latest (hopefully the final) iteration, Naturally there will be only one centre but there will be 5 each of the other two. Making segmented components is usually just a matter of making 6/8/10/12 etc. pieces with the appropriate angle (60°/45°... etc.) and simply gluing them together with butt joints but for gears I need more strength so have designed half-lap joints. It makes the whole thing more complex but vastly superior.

Using Plywood would overcome all the issues of course but has the disadvantage of ‘Appearance’.

The shape of the two different segments can be seen in both plan and elevation in fig-69. They will both be made using CNC and all 5 of each together from two blanks - see Fig-70

. . . . . .

It’s somewhat disturbing to realize that another day has come & gone and nothing material has come out of the workshop. A great deal has been done on the PC of course and all the G-Code has been produced so tomorrow I hope that I might have at least an Escape Wheel blank glued up ready for the outline teeth and spokes to be cut.

 

[Yojevol]

Just spent a few minutes trying to get my head round this. So, if I'm right, you're going to end up with a 10mm thick escape wheel fabricated from 10mm thick pieces joined by halving joints. The only comment I can make is that the outer rim will be made up with a mixture of solid and 2-ply wood. The strength, not that it needs much, will be limited by the single ply sections. At 10mm thick there is plenty of surface area for a simple butt joint which would be stronger than the timber itself, especially with the accurate shaping that your router can provide.
I'm really interested in this section of the work because I'm just beginning to think about making my next clock with solid wood gears. I will have 2 options. (a) similar to your method but with simple segments going right into the centre, or (b) making my own ply from a large amount of veneer which I really ought to find a use for. I'll probably experiment with both.
Brian

 

[J-G]

Not quite right Brian - the Gear will be only 5mm finished - well that is the current plan - after I've made the first I might change that to 6 or 8 but I don't want to go any thicker if I can help it. The 48mm wide Maple at 15mm thick does give me the option of going to 6.5 ish - if the band-saw doesn't wander
This is as much to do with economy as anything - I'd hate to mill away more than half of the board!

I don't believe that the strength will be reduced by the half thickness for the same reason as you cite the glue strength of the butt joint - - - - but I could be wrong

Whatever - it's a great leaning curve both as far as making Gears and the whole concept of CNC machining (which is what this whole exercise is really about).

 

[donwatson]

Thanks again JG. I am afraid you are both beginning to lose me in these details. I need some time to study the PDF.

 

[J-G]

I hope you've had enough time to assimilate those details Don It took me some time to get it clear in my head but once it 'clicked' it became obvious. As you'll see from the next installment (.PDF attached) I've made the first Escape Wheel - based on the half-lapped segments, I've also made some spindles and Pinions.....​

===========================================================================
14th Oct
The larger segment went to plan with no problems but the smaller one gave me a few anxious moments as I watched the 3mm cutter get closer to the clamping screws than I have bargained for! It just cleared the first but partially unscrewed the second. I was ready with the Emergency Stop though and once I’d turned down the head, removing the damage, I could re-secure the blank and continue the G-Code from where it had stopped.

This was caused by my ‘thrift’ since I’d decided to ‘save’ the 50mm of Maple that was left over after cutting the two lengths needed and I’d also put just two clamping holes on the centre-line rather than 6 holes on the edge as I’d had to do with the large blank. You’ll see the difference in Fig-71 & 72.

Fig-71

Fig-72


After separation and a little clean up, I found it a little challenging to get all segments to sit nicely together until I hit upon the idea to screw the centre down. Once I’d done that, the glue-up was straight forward. I was anticipating having to screw 10 blocks in a circle and add wedges to press them all to the centre but eventually was happy with just a panel pin against the end of the long segment. (Fig-73) Whether it will ‘move’ overnight is another matter!

Fig-73 . . . . . .. . .Fig-74

15th Oct
There is a slight ‘bowing’ due no doubt to the fact that I didn’t take enough care after hammering the pins in to also clamp the whole assembly ‘flat’. Nothing that a short time on the Linisher couldn’t overcome but I’ll know for the next one.
Spent the morning re-doing the G-Code for the spokes, Teeth and Pins. The later was of some concern since I want them to be a press fit - though they will be glued as well - and although CNC machining is accurate, the actual diameter of a drilled hole is governed by not only the diameter of the drill used but also by how accurately it is held in the collet. Because I made my own holder for 1/8" shank cutters (Fig-74) I was anticipating some minor ‘wobble’ and when I’m looking at 2mm Ø, a 1.9mm might well cut just a smidge oversize and when I measured the St-Steel pin material at 1.97mm it seemed sensible to do some test holes.

I was pleasantly surprised to find that the very first test hole with a 1.9mm dill proved to be accurate and a good ‘press fit’.

Because the drill was in place, it made sense to keep it there and change the order of operations, so, as you can see in Fig-74, the 30 holes for the Escape Wheel Pins were drilled before the Spokes or Teeth were machined.

The CNC operation went pretty much to plan though there were times when I thought that my 3mm cutter was cutting oversize - it turned out that I had left the offset between roughing and finishing the same so the second cut although being with a smaller cutter only trimmed a little out of the places there the larger cutter couldn’t reach. The re-writing of the G-Code soon sorted that so I now have one of the four Escape Wheels (Fig-75) ready to have the pins inserted. It still needs to have a recess machined to take the spacer between it and the Pinion which will also carry the grub-screw which locks it to the spindle. The centre hole needs to be enlarged as well but these two ops. were always going to be done on the lathe.

I’m quite happy with the thickness at 5mm but because I have the 15mm thick Maple I’m minded to simply cut that down the middle and adjust the depth of the joints once I’ve cleaned up the surface so the other three may well finish up at 6mm or just over. That will just mean that the spacers need to be a bit shorter. The current design is 16mm so there’s enough ‘meat’ in there for the 3mm grub-screw even if it goes down to 14mm.

16th Oct
I have to visit a wood-yard (Whitmore’s Timber - Claybrooke Magna) this afternoon to select a Walnut board for the rest of the Frame and Dial so I started to make some of the spindles, specifically those for the Escape Wheel and 2nd Train since they are both the same length but with a different shoulder to position the gears offset to each other. These need to be concentric so I can’t trust even a newly bored set of soft chuck jaws to hold true when they are reversed. Therefore the only option is to turn them between centres which eliminates all potential concentricity errors. I had all 8 spindles machined to length, centre-drilled each end and one end fitted to the 8 x 4 x 3mm Deep Groove Ball Bearings before going to Whitmore’s.

A vast selection of Superior Grade Black Walnut boards and the yard-man couldn’t have been more helpful in allowing me to see many before selecting what we each considered the best ---- or so I thought!! ---- everything went well and he even pointed me to their ‘offcuts’ section where I found a 1.2m length of 8 x 1 Canadian Hard Maple @ £15. Took a while to deal with paper-work and payment but I was on my way less than 40 minutes after arriving. Only when I got home did I actually measure what I had

Neither he nor I used a tape measure, I just accepted what he said was a 6" board - it turned out to be a 5" board - actually 130mm wide but unfortunately just not wide enough to get my cutting list out of. Another 10mm and I might have been able to re-arrange the cutting up. So I now have to return it on Monday to select another.

So, back to the spindles - mind you, I can at least have a go at the Maple over the weekend to see whether my concerns over the grain direction are really worth the hassle of making 10/11 segment composite Gears.

The Escape and 2nd Train Spindles are now done - Fig-76 shows them comple

te with the bearings.

17th Oct
With the Walnut not available and the spindles made, I thought to make the 8T Pinion which goes on the same spindle as the Escape Wheel.

As all the pinions will come from the same 10mm thick Maple stock, it made sense to set them all out in one drawing so that there would be minimal waste due to having to provide clamping holes, and, after a couple of attempts I finished up with the layout in Fig-77 where I’ve separated the four different Pinions by applying a colour. Combining all four meant that I had to think carefully about the depth of cut when creating the G-Code and the order in which each MOP should be made but that’s nothing new.

I was also aware that these small components also had quite delicate 1mm thick walls which, when assembled (glued in), would be perfectly good but in isolation may well prove an issue so I made the ‘finish’ cut on these parts 0.2mm. The fact that I’m using a 1mm Ø cutter for the finish also had a bearing on that decision!

I’ve learned a little about how Maple handles when cut with an up-cutting end-mill - it leaves ‘whiskers’ which need to be removed by hand, well riffler file - I may have to make another two, if so then I’ll change the cutting order so that the surface is removed after the teeth have been cut, that should eliminate the clean-up, though the reasons for cutting the surface first was to make the depth of cut short enough to be covered by the flute length of the 1mm end mill (and that is still a valid reason) so there may have to be a two stage process where most of the waste is removed leaving a ½mm for a final cut.

Because the Gears are 5mm thick and the 8T ones have a 3mm boss there was at least 2mm and as much as 4mm to clear from the opposite side of the blank so it was always going to be necessary to turn the billet over. When I designed the G-Code, I made a point of creating a Pocket MOP to deal with this but still leave a 1mm bridge between each Gear tooth and the main body. Fig-78 is after the second operation with the gears still held in place but machined to the correct thickness. Most of the Pinions could be broken out but I did need to cut through some of the Maple with a Jeweller’s Piercing Saw.

​

 

[donwatson]

Thanks again JG. This is absorbing reading and a very good example of a true WIP.

 

[J-G]

Thanks for that Don. If I didn't write it up 'on the fly' it would never get written.

Here's a tale of caution that went wrong. I'll add it to the next .PDF but while it's fresh in my memory it's worth noting.

While the Pinions were being cut I got on with making the 1st Train Spindles. They weren't finished yesterday so didn't get a mention but I did continue with them this morning. Again turning between centres and, due to their length and the fact that they have a 3mm Ø section, I had decided that I needed to be cautious about the depth of cut --- that was until it seemed to be going so well that I got 'greedy' D'oh!!

 

[donwatson]

Oooooppppssss

 

[J-G]

18th Oct
While the Denford was cutting the main 72T Drive Gear I re-made the fourth spindle. Before that I took a 310mm length off the Maple board and split that down the middle (ish) - at 150mm+ wide I’ve discovered that there is no way I’ll get three blanks out of the 27mm thickness, particularly since the saw cuts on the face are about a mm deep, so I’m resigned to having to skim at least 2mm off both faces.

As I anticipated, there is some break-away on some of the tips of the teeth due to grain direction but I can’t tell yet whether this will impact upon how the clock ‘works’. At 148mm Ø the 72 tooth Gear (Fig-80) is the largest and is the main drive wheel – ie. it is attached directly to the weight-cord winding bobbin. I had some difficulty with my initial plan to finish cut the teeth with a 1mm Ø end-mill since the flute length on that is only 6mm and this Gear is 8mm wide but to make sure the tooth width is covered I’d written the G-Code to cut 9mm deep.
In the event, I changed that to 6mm and created another MOP to use a point 8mm ‘burr’ for the last 3mm. Once the Gear had been cut, I turned the blank over and cut most of the excess Maple away using the surface MOP, well until the Gear was close to finished thickness - sufficient to break away the waste anyway - the final thicknessing was done on the lathe as was the final sizing of the bore and the clearance recess for the Ratchet.

I’ve included the 8 & 10T Pinions in Fig-80 to show the relative scale of each but it also shows the difference between the ‘white’ Maple I bought a while ago and the Canadian Hard Maple I bought with the Walnut last week.

19th Oct
I cannot praise Whitmore’s Timber highly enough for their service, (their paperwork processing is a different matter!) I returned the 5" wide board this morning and came away with a 9½” wide board as a direct replacement since that was the nearest they had over the 6" I’d paid for. The nett result is that I paid ~£85 cu/ft rather than the ~£140 cu/ft anticipated. With a second visit I also cemented some connections, building a useful relationship for future purchases.

Now I have the Escape Wheel, Pinions and some spindles made, and, after cutting the Maple board for the Main Drive Gear, I have a suitable ‘stick’ from which I can make the spacers. While cutting that up to make blanks, I also made blanks for the Winding Drum because there were oddments of Walnut, Oak and Ash laying about that were of a suitable size so Fig-81 gives you an idea of how these small items start life and Fig-82 shows the three stages that lead to the finished spacers which go between the Escape Wheel and an 8T Pinion.

. . . .

Tomorrow I should start cutting up the Walnut board to make blanks for the Dial but I suspect that I’ll start by finishing the spacers and winding drums.

20th Oct
As anticipated, I pressed on with the Spacers & Winding Drums, what I didn’t expect is that it took me all day!

I’d forgotten that I needed two sets of spacers - one at 16mm & another at 14mm wide - and there is another part (a Boss for the Winding Spindle) which looks similar to a spacer and is on the same page so I made those instead of the second Spacer. When I came to the Winding Drums, I machined the one end and decided that trimming the excess length that was needed as a means to hold the blanks in the lathe chuck would be better removed on the band-saw. Regrettably I was somewhat cavalier (read stupid!) in my regard for that machine and had a bit of a ‘nasty’ resulting in one of the part finished Drums flying off to somewhere in the workshop yet to be discovered, I know it flew past my right shoulder towards the Jewellery bench but I didn't see it land and it could have bounced anywhere.

Having spent far too long searching for the lost part I took the pragmatic path and found another piece of Walnut to start again. The second operation of the Drums wasn’t without it’s ‘challenges’ either. Whilst I do try to be very detailed with my working drawings there are times when I leave out ‘obvious’ or non-critical dimensions and in this case that became a ‘pot-hole’. Working out a dimension for the overall length of the Drum in my head, I failed to take account of the collar thickness so I finished up 3mm short. This made me review the General Assembly drawing - to determine if it really mattered - and since that was only on the PC in my office, I spent some time doing a physical printout -- at 85% size -- on 20 sheets of A4 which needed to be trimmed and stitched together - Fig-84. It was my original intention to pin it up in the workshop but at over a metre wide and high that’s a lot of wall space! though I may be able to fix it to a batten and hang it on the open door. Not that I will need to review it that often - the design is pretty much ‘in my head’ due to the time I spent at the drawing stage but there are over 150 components so the nitty-gritty detail does occasionally get blurred.

. .

In Fig-83 I’ve assembled a selection of the Winding Drums - to show both ends and also with it fitted to the Ratchet - along with the Boss (which will eventually have a brass sleeve) and a Spacer with 8T Pinion attached.

I still need to do further work on the drum to cut the Cord fixing hole and I have to make the matting piece which turns it into a ‘bobbin’, so I won’t get to cut the Walnut up today, and maybe not tomorrow. At least it is acclimatising and could well do with a few more days to settle the MC after being stored outside, under cover but not ‘In Stick’. I don’t have a moisture meter but the board did ‘feel’ somewhat damp so that may well be the reason I seem to be procrastinating.

Can't tell why but the PDF file refuses to show any text if I put two pages together so there are two separate pages to WIP-15.PDF (attached) - I had a similar issue with WIP-7.

 

[donwatson]

Thanks again JG. Looks like you have lots of problems with these pieces but seem to resolve them pretty quickly.

 

[J-G]

I suspect that I don't have any more problems than anyone else but because I'm writing this up as a diary, I record the issues as they arise whereas with a WIP written up after completion the problems get forgotten.

All problems are down to my own ineptitude and I don't find need to hide that since the whole point of a WIP (in my opinion) is to highlight what issues may come up in a 'fore warned is forearmed' attitude.

 

[donwatson]

I agree with sentiment.

 

[J-G]

22nd Oct
I’ve done nothing on the Denford for a couple of days so nothing more to show as far as gears or frame are concerned. I still haven’t started on the Walnut plank - but I have been thinking about the potential issues - mainly due to the problems I had cutting the Maple down the middle of the 27mm wide board.

I’m thinking that I’ll need to break out the Table Saw rather than hope that the Band Saw will behave. That will mean that the kerf will be about 3mm rather than ~1mm so I might have to re-think my cutting regime. Since the board is 9½” wide that won’t matter but it is a large lump and having another pair of hands on it will certainly be an advantage so I’ll probably make a start on it on Sunday when my grandson should be available.

The winding spindles are now made - well the turning part is done - they still have to have the square for the Key and the locking flat milled, and I’ve tentatively assembled the Escape Wheel/Pinion/Spindle together and tested how that aligns with the Latch/Finger/Pendulum as seen in Fig-85. I’ve attached the 1st & 2nd train spindles but the Gears for those aren’t made yet so that is my next task.

There are two gears on both spindles, a 32 & 60T on the 1st and a 64T + 8T Pinion on the 2nd. For a number of reasons I’d cut part of the Maple board down to 132mm wide so could get all the 32T gears out of a 340mm length of that but the 64T needs just a little wider to make sure that there is no ‘trimming’ of some teeth. This made me settle on creating the G-Code for 8 - 32T gears first and I needed to split the Maple down the middle which didn’t go as well as the first time. Hard Maple really is ‘hard’ - beautiful but challenging! Even with the Band-saw tension set for a 1" blade rather than the 3/4" I’m using the cut wandered over 5mm across the 132mm width after 200mm length so I started again from the opposite end but it took some effort to separate the two halves.

I might retrieve the second piece and make two 60T gears from it but first I had to machine the first half to get it near parallel before mounting it on to the Denford table. It still needed a 3mm cut before the surface was clean though. In Fig-86 you can see that it’s finished up at about 9mm thick and it’s interesting to see a dark streak which is totally enclosed - not that it matters, it’s a natural product, and just adds character. The gears will be 5mm thick finished so there is enough timber available to not need a sacrificial sheet and once the gears are cut to 6mm deep the board can be turned over to have the second face machined leaving small tabs holding them in place to be sawn free in a third operation. This may not be possible for the 60T gears I hope to get out of the other board.