CNC Clock Build W.I.P.

[J-G]

Have you learned anything Don? - or have any suggestions regarding my methods?

 

[Cabinetman]

As Don Watson said, I hadn’t seen this before either, strange. So another wooden clock man, well Mr Harrison became rather famous doing it and won the prize for calculating latitude or was it longitude back in the day. And he came from not far from me here.
As the neophyte said totally beyond me but I’m referring to the CNC bit, I can imagine it’s probably the only way to make Clock parts accurately enough, mind you Mr Harrison managed it ha ha. Ian

 

[J-G]

As Don Watson said, I hadn’t seen this before either, strange. So another wooden clock man, well Mr Harrison became rather famous doing it and won the prize for calculating latitude or was it longitude back in the day. And he came from not far from me here.
As the neophyte said totally beyond me but I’m referring to the CNC bit, I can imagine it’s probably the only way to make Clock parts accurately enough, mind you Mr Harrison managed it ha ha. Ian

Longitude it was Ian - thanks for your response though

There have been many superb craftsmen who have done a much better job than I can achieve without the benefit of CNC. I have been inspired by Giovanni De Dondi who built a 7 faced clock over 16 years starting about 1348. This showed not only the time but the position of all the known planets. He even accounted for the apprent retrograde motion of Mercury buy incorporating an elliptical gear! He didn't even have access to gear cutting or threading equipment - all done by hand - one of his gears had 157 teeth, a prime number, cut by hand with a file! (Which he probably made himself)

Then there is the Antikythera mechanism ....... 1st century BC

 

[donwatson]

Thanks JG, I am intrigued by the way you have gone about making these pieces and am too much of an amateur to advise anyone on the use of the CNC router/milling machine

 

[J-G]

I've had a few problems over the past couple of days but I have another two pages done. Much of what is in these is a thought process - working through methodologies of multi-components and work-holding - and there's still a great deal to be done, but I have learned a lot about how different the CNC workflow is compared to piecemeal machining.

 

[J-G]

I've been reviewing the number of [Views] of the attachments and see that the .PDFs seem to be read more than the .PNG files so I've now created .PDFs of installments 4 to 7. I also think that the .PDFs are better quality and easier to read.

 

[donwatson]

Thanks again JG. I find the PDF a good choice as I only need to click on it and it downloads straight into my Downloads box.
I will go and have a look at the latest you have posted, all very fascinating stuff.

 

[J-G]

Thanks for that feedback Don.
I'm also considering the 'Copy & Paste' direct into the Forum as suggested by MikeG as well.

 

[J-G]

Installment 8 attached as both a .PNG and a .PDF.

 

[donwatson]

Thanks JG. Will peruse this later

 

[J-G]

I've created another installment as a .PDF and two .PNG files which I will attach but I'll see what I can do as far as posting the text directly into the forum (Copy & Paste) and insert the images ...................

30th Sept
All seemed to be going well after I’d checked the proximity of the fly cutter to the clamps and finally pushed the [GO] button to skim the surface. I was quite pleased that the blank ‘off the saw’ was 7mm ± 0.2mm and the skimming produced a very good clean face. Doubts began to appear when drilling the pivot holes, some looked to be too close to the front edge and were increased when the first outline cuts didn’t clean up the front edge.

Even with what I thought was careful positioning, the blank was still about 1mm too far back. It was then interesting to see how CamBam had ordered the machining, starting at the left hand end, taking two cuts on the first two Pawls before starting on the 3rd and 4th, returning to the 1st two and so on. I wasn’t too concerned when the first Pawl broke away but when the second not only broke away but also broke across the middle - - - - - - -

The upshot is that I now think that African Blackwood is not the best choice for this component. Yes. it’s hard and will wear well but it seems it is also too brittle. Previously I’ve only used it in ‘slab’ form for the sides of boxes or knobs where, although open grained, is does take a very good finish when ‘filled - sealed and waxed’.

Fig-42 shows the results of this mornings disaster and I’ll now look at my stock of Beech to see if I can do a better job with that.

Not only that, I collected my Brass & Stainless Steel order and after a 20 minute drive home, found that they had supplied a length of 7mm Ø rather than 8mm !!

1st Oct
New layout sorted with just 6 Pawls and two Beech blanks created. I’d decided that I didn’t need a 3mm cutter to cut out the rough outline so designed this batch to use a 1.5mm end-mill. Quickly found out that a 1mm cut at 300mm/min was too much since on the second pass the tool broke - at least that’s the first breakage I’ve had. Changed the G-Code to have 0.5mm depth of cut and 150mm/min feed rate. Took a while to complete but I do have 6 Pawls in Beech which are useable. The second batch are now being machined.

The second Op. went without incident and all 12 Pawls are finished and Fig-43 shows 3 roughly in position around the Ratchet.

It will be a while before I have the part with the Pivot and retaining pins done as that is a 72 tooth Gear which has to be cut from a composite piece of Maple and I haven’t yet even looked at how I’m going to create the blank, never mind cutting the teeth. Though neither should be onerous, just time consuming.

I have made a start on some other components to while away the time taken for the G-Code to run its course.

2nd Oct
A few days ago I prepared some Walnut blanks for the Frame Cross Piece which joins to the Frame Upright shown in Fig-33 - and promptly forgot about them - so today I looked at both how to hold them and skim the surface. The process in CorelDRAW! might be interesting so I’ll try to explain my thinking. Fig-44 shows the component outline within the blank limits. The problem I have to address are how to clamp the blank to the table so that the surface can be ‘skimmed’ without the cutter colliding with clamps. I could, of course, drill & counterbore holes but that would be more wasteful of timber which could be used for other components.
To skim the surface I use a ‘Fly Cutter’ (Fig-12 on WIP-3) which cuts a swathe 50mm Ø so can cover the widest part in one pass. To do so I need to use an [Engrave] MOP which is a single curve.

The red circles on Fig-45 are the plot of where the cutter will sweep and the Grey blocks show where the clamps need to be positioned to avoid the cutter path but they could be turned to give a little more clearance. The Blue crosses show the position of the clamping threads in the table.

This part needs to be machined from both sides so although the centre joint is shown ‘dashed’ to indicate it is on the opposite side, I’ll machine that first - along with the two circular recesses - and use that joint to re-position the blank to cut the end joints. That will be done by screwing a block (green - purple outline) to the table and cutting that to suit the joint, thus fixing the blank in both the X and Y axes.

It might even be best to leave the outline as a Second Op. so that I can skim the second face using the same clamping technique. .... hmmmm..... writing this W.I.P. does have the benefit that some things come to mind sooner through trying to communicate my thoughts. It may well be that I’d come to the same conclusion but only after starting work on a component and then having to change the methodology.

3rd Oct
I didn’t bother with the second ‘skim’ since the first brought the blank thickness within 0.2mm of the target so this morning I got on with making the location jig for the Second Op.. (Fig-46) The first attempt showed that I’d missed a setting in the parameters for the width of the location peg — I’d left it to cut [Outside] (the default) whereas it should have been set to cut [Inside] — there are so many settings so it is easy to miss one if you lose concentration for a moment or two. A ten second job to correct the G-Code but another half hour to cut another blank, drill the fixing holes and mount it on the table before starting to machine it again.

As expected, once I’d corrected the G-Code, the Location Jig was a perfect fit for the previously cut joint so it was a very simple matter to clamp the part machined [Frame B] component ready for the second side joints and holes to be machined and finally the outline cut out.

Fig-47 shows Frame [A] and fitted together.

========================================================================

Well, that went pretty much to plan though I might need to adjust the size of the images.

I'd be grateful for feedback regarding this method of posting -- ie. if it is preferable to the .PDF/.PNG image attachments.

 

[Yojevol]

I found recently that adjusting photo size is easy. Click or tap on the inserted pic and its boundaries will show. Just drag a corner node in to reduce your pic size.
Brian

 

[J-G]

Thanks for that Brian - much better than re-sizing in Photoshop, Deleting the original and Attaching the new version.

It only reduces the display size of course, the file size will remain the same (I imagine).

 

[Yojevol]

@J-G I've just been looking through your posts on the CNC discussion thread. You mention that your bearing design is still under consideration. Having been through the mill on this subject in recent months, I thought I'd put up a few thoughts.
My first clock attempt was Clayton Bower's Simplicity which, like many designs, relies on a simple hole in the gearwheel running on a brass shaft. In order to keep the wheel running true, ie, minimal wobble, a good length of bearing is required. This is often achieved by adding a boss to the basic gear.
For my second effort I have designed it to be as narrow (front to back) as possible which means no added bosses. I thought little deep groove ball bearings would be the way to go; mounted directly into my 6mm ply gears. However it turns out that these bearings have a considerable amount of angular play, so they are not the answer to wobbles. My next try was with needle bearings but on these small sizes they're not much better than ball bearings.
The solution I have ended up with is using 2 bearings ganged together which increases the overall thickness from 6 to 8mm - I need a mm or 2 each side for clearance anyway. Where a gear has to be fixed to its shaft it is often possible to fit 2 well spaced bearings and that's fine as long as I can fix the shaft at a true right angle (that can be a challenge).
I would certainly recommend doing a few experiments before committing yourself to cutting your nice timbers in anger. You may well be in a better situation using quality wood rather than ply. Also your CNC router may give better quality holes compared with my ply drilling methods.
My next project is probably going to be a repeat of my No.2 with a trial use of lignum vitae (just bought an old bowl) as bearing material.
I'm just thinking of doing a thread on the present project - taking it through the final assembly stage. Here's a little spoiler
Brian

 

[J-G]

Thanks for that feedback Brian. I'm sorry to have given you the impression that bearings are are still under consideration, those details were sorted well before I started selecting timber to machine.

All of my gears will be fixed to shafts via a grub-screw clamping on to a milled flat. The bearings will be at the ends of each shaft - though in three cases the shaft will go through at one end. Most bearings (all ordered and delivered) are Deep Groove but two are Needle Roller. I've used both in the past without having any issue with 'wobble'.

Here's a drawing of the 1st Train Spindle which also carries the Hands - I've added the bearings (Purple/Silver) to show where they sit. [Edit] - now added the location of the gears which are assembled with a spacer which is glued in (red line = glue line).

Certainly solid hardwood and CNC gives me an advantage over Plywood and a Pillar drill but knowing the limits of your tools is part of any engineering equation

Good luck with the 'Bowling Ball' - I was given two a while back but every attempt to get anything useful out of them has been thwarted by bits falling off due to shakes. My Grandson did make use of a small block to hold some Silver tubes that he needed to file to length but even that broke apart - fortunately it gave him the chance to push the tubes out when finished!

I did look at your 'spoiler' but [DropBox] has a mind of its own so I could only run it once.

 

[MikeK]

Well, that went pretty much to plan though I might need to adjust the size of the images.

I'd be grateful for feedback regarding this method of posting -- ie. if it is preferable to the .PDF/.PNG image attachments.

Don't bother resizing images, unless they are larger than 10MB. You can upload PNG and JPG images in their full-size format in line with your text. The XenForo software will automatically resize the image for viewing based on the device. Clicking on the embedded image will display it in its full size.

It's not possible to embed PDF files, but they can be attached to the post as you have been doing.

 

[J-G]

Thanks Mike. It was just that when the images came in they seemed larger than I'd created --- I hadn't considered what you have intimated in-as-much-as the forum can be viewed on devices other than a PC.

There was/is an option to embed a Thumbnail or Full Image - I've selected Full but Thumbnail might be more efficient, if click on the thumbnail displays the full version.

 

[MikeK]

Thanks Mike. It was just that when the images came in they seemed larger than I'd created --- I hadn't considered what you have intimated in-as-much-as the forum can be viewed on devices other than a PC.

There was/is an option to embed a Thumbnail or Full Image - I've selected Full but Thumbnail might be more efficient, if click on the thumbnail displays the full version.

Please select Full. This will display the image in the post without any action required by the viewer. If you select Thumbnail, then the viewer must click on the image to display it in full size.

 

[donwatson]

Thanks again JG. This is fascinating stuff for me and now that Yojevol has joined the discussion it helps me to rethink my position regarding making wooden geared clocks. I was beginning to think it was not for me but I now feel I should give it another go. (Whenever time permits).
I find the PDF is a good way to store the articles on my computer, a click and it downloads into my 'Clocks' folder. But I also like the way you have just presented it on the forum, it is clear and I can understand what is going on at a glance. I like reading it here but I like a copy on my computer that I can peruse at my leisure without having to log on to any forums, though I have to admit I wouldn't be here without the forum. Hope this helps.
EDIT Hi JG, I have been looking through this fascinating article and can only say it has re-kindled my interest in making a wooden geared clock. I have found a piece at the end of WIP 7B where you designed the composite gear and made it from 5 pieces. It puzzled me for a bit as I would have thought 4 or even better 6 pieces would have been easier to work with (almost mental arithmetic) but then thought it would be in keeping with the statement you made at the outset about using the 'Releaux Polygon' so that sort of solved that.
At the end of this same article, in the downloaded PDF, the last line is missing ??
Thanks again for this fascinating insight.
take care
Don W

 

[J-G]

Thanks Don - very useful feedback - I'll certainly continue with the .PDFs