CNC Clock Build W.I.P.

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
17th Nov
I left the Denford working away last night and was disappointed to find yet another broken burr and an unfinished gear when I went in the workshop this morning. I can’t see why three new burrs from the same batch have succumbed to the smaller stresses - after I reduced feed, increased speed and halved depth of cut but it seems prudent to change tactics. To that end I’ve taken a day off to ponder the options.

Fig-120.png
It’s too late now to do anything with the gear in Fig-120 since I’ve taken it off the Denford table and the chance of re-mounting in the identical location within the 0.01mm tolerance that would be useful is impossible so I’ll have to finish the last 2mm of each tooth by hand.
Fig-121.png

I have mounted the next blank and run the G-Code as far as roughing out the teeth. Whilst that was running I looked at what cutters I had that would handle the 9mm depth. All the Burrs will, but the smallest I have (above the 0.8mm) is 1.8 and the small end mills won’t go deeper than 7mm.
Not to be defeated, I decided to modify a 1.5mm end mill by grinding away part of the shank such that it wouldn’t foul the top of the teeth - Fig-121 shows the result.

Tomorrow I’ll run the G-Code to finish the teeth with the 1mm end mill which can get down to 7mm and create new code to use the 1.5mm cutter for the last 2mm. Because the 1.5mm cutter won’t be able to get quite as close to the ideal tooth form, it will mean that I’ll still have to do some ‘fettling’ on the teeth once I’ve cleaned the face of the gear to 8mm thick - but I won’t have run out of the 0.8mm burrs!!

17th Nov
Sleeping on a problem - I often find - has a beneficial effect and, after I’d set the G-Code running using the 1.5mm modified cutter, I had a look at what options my mind had brought to the fore overnight.

My concern was that the gear tooth form could not be followed closely with a 2mm Ø cutter – ie. that there would be potential for binding because the ‘root’ hadn’t been fully cleaned out – this was only from a ‘thought process’ rather than empirical study. So, - I’ve now done a drawing of the tooth form at a 10x enlargement and along that plotted the path of the centre-line of the cutters at 2.0, 1.5, 1.0 & 0.8mm Ø.

Fig-122.png
As you can see from the drawing which is Fig-122, I was surprised to find that my concerns were totally unfounded. Even a 2mm cutter will reach into the deepest recesses of the root so all my faffing about - read that as breaking tooling!! - over the past three days has been without benefit :( Unless of course you consider that the learning curve has its own benefit!

I think that I’ll still use the 1.5mm cutter (especially since I’ve already increased its reach) because it is an ‘Up-Cut End Mill rather than a ‘Burr’ so should leave a cleaner finish.

I’ve just checked and even a 3mm Ø cutter would still cut down to the root correctly but a 1/8" (3.175mm) would not, and I’ve only been able to source Up-Cut end mills at 1, 1.5 and 3.175mm Ø (at sensible prices anyway).

18th Nov
The last 72T Main Drive Gear did get completed using the modified 1.5mm end mill and all four were finished to 8mm thick on the lathe where I also finish bored out the centre hold to fit the p
Fig-123.png
ositioning Boss and the Ratchet clearance recess. All four are assembled (Fig-123) on their spindles along with the Ratchet and Winding Drum but without the Pawls or Pawl Pins - the Pins will be a press fit and I only want to do that operation once so it will be left until after I’ve applied the ‘finish’ which will be Sanding Sealer and MC Wax.

I also found time to cut the slot in the Lower Frame Spacer to accommodate the Frame Brace.

Initially, I thought to cut the slot on the Milling machine but soon realized that it would be both quicker and more accurate to do it on the Denford - primarily due to the difficulty of getting the parts absolutely level in the Milling Vice. The Denford table is, by design, flat and
Fig-124.png
level so setting up a clamp to accurately position the 16mm Ø spacers was a few moments work and the actual machining would only take 2 minutes at most — it just meant that I had to spend 30 minutes or so drawing the tool paths and creating the G-Code. I suspect that it would have taken an hour to do the job on the Mill with the prospect of error.

The M5 threaded inserts which will eventually be screwed into the Front Frame have all been fitted to the clamping spindles so that I know that I have sufficient and don’t need to order more stock. Fig-124 has one of the rear frames loose assembled with the Frame Brace, Spacer & Clamp spindle.

19th Nov
Back to CNC and Gears - When I made the 32T Gears I was still drilling and counter-boring holes in the basic stock to machine the surface down close to the finished thickness but I’ve learned from making the 72T gears that putting the part finished Gear on the lathe to skim to finished thickness is easier to do - both as far as holding and machining is concerned. For the 30T gears I made today I didn’t even cut the 400 x 64 x 10 Maple blank down to the 297mm length that I needed to cover the four so there was room to fit a 10T Pinion on as well – I needed another due to knocking a part assembled unit over and after a week of hunting still can’t find the Pinion that was in the assembly. Now I’ve made another I’m sure it will turn up!

Up to now I’ve left the ‘waste’ at the thickness of the blank and used a Jeweller’s saw to cut the components out, thinking that I might want to use the ‘waste’ for other projects but I now have boxes full of off-cuts of Maple and Walnut and am less of a scrooge :) so when I prepared the drawing for these gears I planned in an area to be machined away so that the gears could be cut off and held in the lathe chuck.
Fig-125.png

Fig-126.png

In Fig-125 you can see the gears with the waste removed and Fig-126 after they’ve been separated on the band-saw ready for finishing on the lathe.

Because the width of the blank was very close to the O/D of the gears, I had to be very precise with the positioning in the Y axis such that even adding strips of paper/thin card made a difference.
 

Attachments

  • WIP-25.pdf
    1.7 MB · Views: 2
20th Nov
This is a day I will remember!!
Fig-127.png

I’d determined to give the Denford something to do while I set about finishing the 30T gears and rather than start on the Dial – for which I have the blanks semi-prepared – the 64T & 60T gears looked to be the best bet.

My work-flow now starts with determining what size the blank material is so naturally I needed to look at the Maple plank I have to make sure that I could cut three slices at 7+mm thick. I knew it had some ‘edge damage’ (see Fig-127) but imagine my dismay at realizing that when I cut a piece off the one side, to make the Frame Spacers, the board finished up at 130mm and to make the 64T Gears I need a clear 132mm !!

Clearly a re-think was needed so my ‘fall-back’ position turned out to be making up composite boards using the 64mm x 10mm thick Maple offcuts which I rebated along the 410mm length to make half-lap joints. As I was preparing a suitable clamping jig for the glue-up I had a phone call from my Grandson (who lives with me) enquiring if I was home. On hearing that I was, he suggested that I not venture out since he had just been tested Positive for COVID-19 :(

I knew I should have gone shopping yesterday but my ‘needed’ list was small. The rest of the day was taken up with finding out what I needed to do as far as getting a test was concerned and learning about the horrors of on-line-shopping!!

As I said - a day to remember!

21st Nov
Back to the Maple glue-up and the 64T gears. I finished the clamping jig and glued up the three lengths to give me a 160mm x 410mm board (Fig-128 is the jig with the Maple under pressure) from which I can cut three Gears. The Denford has a 400mm ‘soft’ limit on the X axis but by judicious testing this morning I’ve found that I can push it to +205 so I do have a little more leeway than I thought.

Fig-128.png
At 162mm the board is just a little too wide to allow me to clamp it on the table using the eccentric jaw so I’m going to return to clamping by screwing down. I don’t have to skim the surface so there is no need to sink the screws and the heads will be avoided by using a 6mm Ø cutter to cut outside the red line in Fig-129 to clear the area around each gear to ease the separation on the band-saw.
Fig-129.png




The space between the gears should be cleared by the smaller cutters when they are forming the teeth.

The brown bands in Fig-129 show the position of the half-laps.
Fig-131.png

That went pretty much to plan though I did stop the first clearance profile when it got to the bottom of the centre Gear because it was too low. A small adjustment to the position of the whole drawing in CamBam and a re-generation of the G-Code soon corrected that.

Fig-130.png

I surmised that the clearance space between each gear would be removed by the 2mm cutter but you can see in Fig-131 that was not the case. Not that it matters :)
 

Attachments

  • WIP-26.pdf
    1.1 MB · Views: 1
Thanks Don, since I don't have any symptoms (neither does Steven) I can't get a free test -- I'm sure that I could get a test at one of the 'walk in' centres I've been told about but that would mean breaking my isolation which sort of defeats the object.

Combine that with the fact that any test is like an MOT - it is only valid at time of issue - and there is no guarantee that, having been tested, one cannot pick up an infection on leaving the test centre - or at any time on the journey back to isolation.

Steven has been tested every week for the past month since he is working in a food distribution centre and the company are testing every member of staff regularly. Upon testing positive he was immediately escorted off the premises and bundled into a taxi !
 
22nd Nov
A busy and productive day which started with making another composite board as a blank for the last 64T and the first two 60T gears. Naturally, that had to be left on the clamps until the glue had cured so there was plenty of time to create the new drawing and G-Code. I soon realized that I’d forgotten to measure the width of the new board (the Maple off-cuts are not all the same width) so although I could place the outlines of the three gears roughly in position, I would have to wait until I could release it from the clamps before finalising the G-Code. Just as with the previous board, I’d decided to cut a 6mm wide clearance ‘slot’ but, knowing the mis-alignment issues I’d had, thought to create a ‘test cut’ only ½mm deep to make sure that I didn’t need to stop the run and re-align after a deeper cut.

As expected, cutting the three 64T Gears out on the band-saw was easy, and left me with a clear means to grip them in Beech Jaws - Fig-132 is a photo taken after I’d started to cut away the excess – ie. I nearly forgot! – You can see that with the blank at 10mm thick and the gear outline being 8mm deep there is a 2mm ‘slab’ that has to be removed
Fig-132.png
just to get the waste between the spokes out. Ultimately the thickness has to be reduced to 5mm and because these gears are cut from a half-lapped composite board that has to be done from both sides.
In Fig-133 you can see the 64T Gear still in the Beech jaws at the finished 5mm thick. Similarly, Fig-134 is the 30T Gear held in a smaller recess in the jaws.
Fig-133.png
Fig-134.png


26th Nov
Although there has been a 4 day gap since my last comment, I’ve not been idle, it’s a combination of other commitments and the fact that nothing had gone wrong and worthy of reporting – or so I thought!

Much of the time has been spent on repeating many of the tasks already mentioned but also further work on the rough-sawn Dial Blanks. Since these blanks will need work on both sides, it is imperative that they can be positioned accurately when ‘flipped over’ so the first thing I had to do was make a positioning jig - nothing extraordinary just a piece of 10mm ply screwed to the table and machined in position to provide constant X and Y axis end stops. There are 20 blanks in total and cleaning the surface by taking 0.5mm skims off each side until such time that all the saw marks were removed simply took what seemed forever.

Once they were all clean and finished at 6mm thickness, the ends also needed to be machined square. I had determined that they needed to be 233mm long but due to a couple not cleaning up at that they finished at 230mm – no real problem I thought, a little ‘tight’ but judicious positioning of the outline should sort that out. To make absolutely sure of that I set about printing ‘life size’ images of the five Dial sections so that I could easily check how much ‘adjustment’ might be needed.
Fig-135.png


Drawing sections A & B didn’t cause me any concern but I was somewhat miffed when I looked at the overall length of section C and noticed that it was 242mm long...... again ‘Measure twice, cut once’ comes to mind but in this case that should be ‘check all components twice’! Section D & E do not present a problem but another ‘annoyance’ is that I could, with a little forethought, easily have made the blank for ‘E’ shorter and ‘C’ longer. I now have to start from scratch to make 4 blanks for section C. :(
 

Attachments

  • WIP-27.pdf
    1.2 MB · Views: 3
I think you are making an outstanding job of this and can only commend your perseverence and your knowledge of the subject. Thanks again for the info.
 
Thanks Don -- you're not the first to reference my perseverance :) I seldom give up on any problem without a positive result as many of my computing customers will attest.

It's been a few days since I posted an update so it seems reasonable to confirm that it continues at 26th Nov.
==============================================================================

That was not as onerous as I thought - I still have Walnut in the original plank, the table saw is still set up (it often resides tucked under a bench-top since I don’t use it that often) and the Denford still has the Dial Blank Jig fitted.

I took the opportunity to re-surface all of the Dial blanks down to 5.5mm, leaving just 0.5mm excess over the finished size so I don’t have to mess about making the depth of the joints cut from opposite sides different, I’ll make them both 2.75mm. This will also help to make sure that when I cut the slots for the Hour Battens and holes for the Minute Dots I can afford to go 2.5mm deep to make sure that they won’t be trimmed out of existence when I do the final surface skim.

27th Nov
I decided that it would be good to have the Battens available before I start machining the Dial parts. I had initially thought to make the Battens on the Mill but sanity prevailed and I realized that it made much more sense to cut them on the Denford and once I started the drawing saw that making two together - end to end - would also be more efficient. Fig-136 is a drawing of the Dial Section A
Fig-136.png
with Maple Hour Battens & Minute Dots to show the potential problems that I’m likely to encounter.

You’ll notice that the Batten on the right is very close to the joint. This particular joint is in fact at the 12 o’clock position which has two Battens - one each side of the joint line - with only 3mm between the battens at the closest point, so having the Batten in place before I cut the joint will afford some protection to the Walnut.

With this in mind I created a drawing of the Battens as in Fig-137 after preparing a small piece of Maple - 150mm x 48mm - and skimming the surface down to 3mm thick. This would be sufficient to make 24 Battens (in case there were ‘failures’) and I only need 16 . . . . . well that’s what I thought . . . . What I actually need is 16 for each Dial - so in reality I need 64!
Fig-137.png
When I created the G-Code, I set ‘holding tabs’ in the centre top and bottom of each Batten pair thinking that this would be sufficient to hold them all in place and would be where they would be separated. The reality was that cutting right through the 3mm caused some ‘chatter’ at the bottom right of the first column but that only lost one so I left the Denford to com
Fig-138.png
plete the second column. Coming back a while later I wasn’t pleased to see the mess in Fig-138. At least I didn’t come back to find a broken cutter! That was very welcome since it is a brand new 1.5mm Ø Up-Cut router bit with a flute length of 12mm which I found along with a 4mm x 22mm flute and a 2mm x 22mm flute at reasonably sensible prices. I will be able to rescue 16 Battens but of course I need to make another 48. For these I’ll reduce the depth to 2.8mm, that will be plenty to fill the 2.5mm deep slots I have designed into the G-Code for the Dial Sections though.

That did not go well!!. . . . . When I saw the first cut, I thought it was a little wide but let the process continue. It was a fortuitous that I kept a watching brief while doing other work because it was suddenly obvious that the X axis had stopped working again!! I was able to stop the run before too much damage was done but then noticed that my initial suspicion that the cut was wider than the 2mm that I expected from a 2mm End Mill was true. In fact the brand new 2mm cutter actually cut a 2.6mm wide swathe. This meant that the Battens were in fact not 4mm wide as designed and drawn, but just over 3mm. I can only surmise that the length of the cutter (22mm) means that at 15000rpm there is some ‘whip’ which is somewhat disconcerting but may only be a problem when the depth of cut is small - in this case it was 0.5mm.

I’ve now re-visited the CamBam file and modified it to specify my existing short 1.5mm end mill and am in the process of cutting the second batch. I can recover half of the first 24 Batten blank due to the way that CamBam
Fig-139.png
organized the cutting order (Fig-139) but I’ll still need to prepare another thin Maple blank to make another batch.


28th Nov
Fig-140.png

At last I have successfully cut a sheet of 28 - increased from 24 because with a 1.5mm cutter it was possible to get another row in - and another 14 from the first blank.

D’oh! - - - The astute among you will notice that when I dressed this morning I put my ‘Stupid’ head on :( Though that might be true of yesterday as well! I’ve mentioned making batches of 24 and 28 Battens, but if you look at Fig-140 carefully you’ll see that there are 56 Battens so with the extra blanks I’ve made, and without counting my first two failed attempts, I’ve actually made 128 - and taking account of the fact that I’ve cleaned off the ‘Holding Tabs’ on the linisher by hand so there may well be some variance in the finished width I can at least select the best when I come to fit them into the Dial. I also sat and watched as my Fly-cutter ran into the clamps :( I simply thought “it’ll change direction before it reaches those” - - - - Oh no it didn’t.

I can now prepare the G-Code for the Dial sections so I’m starting with ‘C’, which is at the bottom and has 4 Battens, 10 Minute Dots & two joints all cut from the front. The only thing it has on the back is a hole for the Dial Locating Peg.

Having been caught out by the new 2mm End Mill cutting at 2.6mm I thought it wise to do a test cut with the 4mm cutter which came from the same source. Slightly annoyed that I missed the part in the description that said 4mm shank - I had expected either 1/8" or 6mm - it is 4mm, and I don’t have a 4mm collet! One of the advantages of having a reasonably equipped workshop is that I can make tools, jigs and fixtures pretty much as and when needed, so 15 minutes on the lathe and mill and I now do have a holder for 4mm shank tools to fit a ½” collet. I’m sure that is running ‘true’ but the test cut showed that the 4mm End Mill cuts a slot 4.3mm wide - good job I checked!

This means a re-think as far as the G-Code is concerned so I’ll now use a 1/8" cutter and an inside profile MOP rather than an Engrave MOP - I now have to do a test to make sure the 1/8" cutter rings true.

29th Nov
I do hate it when ‘tests’ are inconclusive. A slot drawn at 4mm wide cut with my 1/8" (3.175mm) End Mill set at 3.175mm for a profile MOP finished up at 4.3mm wide but a drilled hole was 3.2mm Ø!! A hole drilled in the same way with the new 4mm End Mill was 4.6mm. The next test was a 4mm slot cut with the tool diameter set at 2.6mm which produced a slot just a smidgen over 4mm. Thinking outside the box I then mounted the 4mmØ Single Flute TCT cutter which I’ve had for some time and did a test drill with that - - - Spot on at 4mm! I didn’t bother with a test ‘slot’, I’m happy to go with the 3.175mm mill being called 2.6mm.

Fig-141.png
It now feels as though I’m moving forward again since I have the section C of all four Dials cut out and with Battens, Minute Dot holes and Joints all cut. Fig-141 shows the Maple Battens in place. These aren’t glued in yet but I will do that on the A & E sections before I cut the joints for the reasons I mentioned on the 27th.

I pressed on with Section D which needed a joint cut on the second side and this did cause me a small problem - it seems that every new component raises yet another ‘issue’ that I hadn’t considered - naturally, cutting right through a blank, even though I am putting in ‘Holding Tabs’ means that the blank is not as solid, so when I turned it over to cut the second joint the grip afforded by the clamping was no where near as high as it was for the solid blank when I cut the first joint. The first time I noticed that
Fig-142.png
this was an issue was at the first cut at 1.3mm deep and the blank just moved with the cutter in the X axis! I was able to abort the run and clamp both ends but I still reduced the cut to 0.4mm - it just took that much longer. By now I had glued the Battens in so I had to cut them back to level on the Linisher otherwise the depth of the second joint couldn’t have been controlled.

I did get all four Dial D sections completed before Strictly and F1 needed my attention though.

30th Nov
The second operation clamping issue made me re-think how I organized the order of operations for the next part - Section B - and I realized that there was no reason that I couldn’t cut the second joint before cutting the outline, so that is how I organized the G-Code for this and all subsequent sections.

That modification to the cutting order was a definite benefit and I now have four of the five sections complete - well as far as the machining operations are concerned, I still have to break them out of the blanks and clean up the joints but hopefully tomorrow will see the Dial ready for glue-up or at least I should be in a position to set up a glue-up jig.

1st Dec
An auspicious day since I entered this world 79 years ago - how time has flown and what changes I’ve seen – or not ?? in 1941 death was all around but so it is today - very different reasons but the bottom line is the same.

I have now finished the last section of the Dial but the manual clean up and final fitting of all the joints - just niggling small (less than ½mm) adjustments to the length of the half-laps so that they ‘nestle’ into each other - on only one Dial has taken me most of the afternoon. Fig-143 is the first Dial simply laid out without glue. It was very satisfying to measure the width and height at within a millimetre of the designed 340mm, by the time it is glued up I anticipate that it will be spot on.

Fig-143.png
 

Attachments

  • WIP-28.pdf
    2 MB · Views: 2
2nd Dec

Fig-144.png
As I hoped, I set up a jig to position and clamp down the 5 sections of each dial while the glue cures. It was heartening to see that the glue just squeezed out of the joints when I gently tapped up the wedges going around them 3 or 4 times. The image at Fig-144 is the first one and releasing that was a little stressful because some glue had seeped out of the back of each joint so I had to apply a gentle tap on a chisel to release it. For the second Dial I added a piece of paper under each joint!

The first Dial has now had the Minute Dots added not as tedious as I had been thinking. Normally I get bored when I have to make more than 4 of the same part and this needed 48 but in fact turning a short length of Maple down from 6mm square to 4mm Ø and cutting of 3mm long lengths went remarkably well. As you can see in Fig-145 the curved joint at the 5 o’clock position is ‘tight’ (as are the others) and the Dots are glued in and stand just a little proud. They will be sanded back flat and level with the Walnut on the Mill using a 50mm Ø sanding pad and then finished with the Makita ROS.
Fig-145.png

I’m not sure that there is any benefit in leaving the Denford to get on with running a G-Code file while I get on with other work.:unsure:

This has been my methodology for some time and I’ve been caught out by something going wrong and destroying the CNC work because I wasn’t watching its every move and hitting the emergency [STOP] as soon as any error occurred.

This was brought home today when the blank for the two Escape Wheels released itself from the clamps allowing movement in the X axis. I was totally unaware that there was anything amiss until after the run was complete and I returned with the intention of changing the tool and setting up the next operation. The first thing I noticed was that the 2mm burr was broken and then I saw the mess that had been made - Fig-146 - I suspect that the Burr had broken because it was being forced to take a deep cut due to the movement of the blank which I suspect was caused by it getting hot and warping because of
Fig-146.png
the moisture content. When I took it off the Denford I could see that it had ‘curled up’ about 3mm - and when you consider that it is only 8mm thick that is a serious amount and certainly enough to release it from the clamp. Basically, this is down to my reliance upon the eccentric clamp rather than drilling clamping holes and screwing the blank down.

4th Dec
To make the blank for the last two Escape Wheels I used up the last stock of my 10mm thick Maple offcuts but fortunately I do still have the Hard Maple board and it is 130mm wide so sufficient to cover the 120mm Ø needed. The only ‘issue’ being that the board is 27mm thick so I had to cut a 10mm thick slice (or rather two slices) off and skim them down to 8mm before running the G-Code again.

Strangely, I didn’t learn from my previous mistake :mad: and still didn’t screw the blank down, nor even clamp the corners, so would you believe that I had a similar failure! - again, for exactly the same reason :(

A third slice off the Maple board and this time I did clamp them down at each corner and eventually got the last two Escape Wheels completed - well, as far as CNC machining is concerned - I still have to mount them on the lathe to trim the faces to finished thickness so there is still an opportunity to scrap them :)

At least I did finish the Dial Minute Dots as my multi-tasking operation, so the four Dials are ready for finish sanding
 

Attachments

  • WIP-29.pdf
    1.3 MB · Views: 5
5th Dec
With the last two Escape Wheels completed I think that all the CNC work is now done so I need a time of ‘reflection’ and checking all the components do fit together. I still have some lathe work to do on the 64T gears, Escape Wheels & Minute/Hour Train since they aren’t yet mounted on their spindles.

6th Dec
I discovered today that not only had I not made the short spindle for the Minute/Hour train, I hadn’t even done the engineering drawing for it! That wasn’t such an issue really because it’s a very small part and it took only 10 minutes to sort out. Fitting the Spindles to the Escape Wheel and 64T Gears proved interesting. The spacers had been made 3 or 4 weeks ago so naturally there had been some movement in the timber and I had to Ream the bores out again and drill & tap the M3 holes which clamp the Gear assemblies to the Spindles. It’s amazing how long it takes to fit components like this together but at the end of the day I did at least have the gears mounted on all the spindles.

7th Dec
Fig-147.png

The first thing I did this morning was to see whether the gears on the three main spindles would mesh together. The rear frame was already fitted together but not glued and I haven’t yet fixed the pinions absolutely to the gears because I want to be able to finish sand the surface and apply sanding sealer/MC Wax before final assembly but there is sufficient ‘interference fit’ to see whether rotating the Escape Wheel causes the whole train to rotate.

It was encouraging to find that I could ‘spin’ the Escape Wheel and the whole train moved. The problem manifested itself when I then fitted the front frame which forced the spindles to be parallel which proved that there is insufficient clearance between the Gear and Pinion mesh. I could rotate the Escape Wheel about a quarter turn which means that the error is not great but still unacceptable. To some extent this was to be expected because I tend to work (draw) to an ‘ideal’ size.

There are a number of ways to effect a solution;

1 - re-cut the location of the bearing seats in the frame - 1mm each would be ample.

2 - Make an abrasive stick and ‘file’ each tooth - or make a jig to do the same thing on the mill using a burr.

3 - re-make the pinions using an effective diameter 1mm smaller.

The first option may well affect the mesh between the main Driving Wheel and the 32T Gear on the centre spindle which would complicate matters further.

Nº 2 would be tedious and liable to make each tooth different.

Fig-148.png
The third seems to be the best option and would also give me a chance to correct the small ‘irritations’ in the way that the CNC process cut the ‘hubs’ which fit in the gears. Fig-148 shows the hub damage due to the method I used to cut them. The Pinion on the right is how they should be but the one on the left shows that the thin wall has broken away - this is due to cutting a circle with an end-mill and proves that it is important to use CNC where the benefits are to be
Fig-149.png
had but not just because it is possible to do the job that way.

With hindsight, I would now just CNC the pinion leaves (teeth) and then turn and bore the hubs on the lathe holding the Pinion in a four jaw chuck.

The blue line in Fig-149 is the original outline/profile of the 8T Pinion and the Red line is the re-worked profile. A two minute job in CorelDRAW!

I’ve created the new DFX & G-Code files to make 8 Pinions and found out why there is a ‘bind’ - I’d created the G-Code for the gear outline ’finishing cut’ specifying a 1.5mm cutter but the cutter I fitted is 1mm Ø !!

8th Dec
Making the 8 Pinions went pretty much without a hitch but I then considered how I was going to hold them in the lathe to both remove the excess material from the back but also turn the hub & bore.

I’ve mentioned previously that I have what I’ve termed ‘Beech Jaws’ to hold
Fig-150.png
the larger gears. These are simply blocks of Beech which I’d previously made to fit on the three jaw engineering chucks that I have. They are fixed to ‘Soft’ jaws which I’ve drilled & tapped. For Fig-150 I’ve removed one of the Beech Jaws so that you can see exactly how this works.

I can prepare the jaws to hold any component truly concentric by first clamping a small piece of steel in the metal jaws and then bore the Beech close to the diameter of the new part that I need to hold. I can also simply glue a new block of wood to the Beech - the triangular blocks are in fact Maple that came out of the gears - and just re-cut them as often as necessary.

This methodology is very good but has a limitation due to being only 3 jaw which is generally no problem when needing to hold objects which are already ‘round’ but not when you start with a square. An 8 tooth gear/pinion is effectively a square component so could not be held easily - even if I were to glue new blocks to make the grip smaller.

I do also have a 4 jaw self-centering chuck with ‘carriers’ - for which I also have a number of jaws but they are all ‘hard’ so can’t be machined for absolute concentricity. Notwithstanding that limitation, having made Beech Jaws for the 3 Jaw chuck I saw no reason that I couldn’t do the same for the 4 Jaw. The small irritation that had stopped me doing this previously was the fact that the carriers on the 4 jaw have a circular location ridge which would be difficult if not impossible to cut on the mill without de-mounting the vice and fitting a dividing head but now I have a CNC machine those constraints vanish.
Fig-151.png
It still took me all morning to design, draw, G-Code and make 4 Ash jaws but the result will be a massive benefit to future projects. I drilled & counter-bored holes for both screw holes in each jaw but the next time I make them I’ll
Fig-152.png
only use the outermost one. This will allow me to hold much smaller objects. . . . . . I could remove the innermost screws and glue a new block in place of course but I have already finished the 8T Pinions and fitted them to the spindles. The jaws - along with the Pinion blank and finished Pinion are in Fig-151 & 2.
 

Attachments

  • WIP-30.pdf
    1.5 MB · Views: 3
Last edited:
Well, this CONTINUES to be what I would describe as being a superb WIP J-G. Again thanks for all your effort.

It's been a while since I've been here 'cos I'm myself working on a lengthy WIP about a child's pedal car toy. It's turning out MUCH longer than I expected (I'm up to Part Ten and reckon there will be about 15 Parts when finished) with lots of words and lots of photos + a few hand-drawn diags.

If you remember a comment I made earlier (Oct/Nov time), I've definitely given up using MS Word (except for initial drafts from "blank paper" which then get "polished" on the actual Forum page). That takes time - and BTW, you were dead right, whatever it is that MS Word does to embedded pix (I'm using .jpg format) the UKW Forum software does NOT like it! So it's loading photos direct on to the Forum from my PC hard drive (I don't do photo hosting).

All of which has nothing to do with your EXCELLENT WIP. It's too late for me and this toy car, but next time ("Next time????? I haven't even finished the current one yet!!!") I shall definitely try the idea of your "planning/diary" approach - that would have saved me from a couple of problems with my "mods with un-for-seen consequences"!!!

BTW, I might have missed it, but I saw somewhere above that someone in your family had tested Covid positive but didn't see anything else. I do hope you and all yours are OK.

Thanks again and if it doesn't sound too patronising" keep up the good work.
 
Well, this CONTINUES to be what I would describe as being a superb WIP J-G. Again thanks for all your effort.
You're too kind AES.

AES said:
It's been a while since I've been here 'cos I'm myself working on a lengthy WIP about a child's pedal car toy. It's turning out MUCH longer than I expected (I'm up to Part Ten and reckon there will be about 15 Parts when finished) with lots of words and lots of photos + a few hand-drawn diags.
These things always take longer than expected :)

AES said:
If you remember a comment I made earlier (Oct/Nov time), I've definitely given up using MS Word (except for initial drafts from "blank paper" which then get "polished" on the actual Forum page). That takes time - and BTW, you were dead right, whatever it is that MS Word does to embedded pix (I'm using .jpg format) the UKW Forum software does NOT like it! So it's loading photos direct on to the Forum from my PC hard drive (I don't do photo hosting).
MS Word converts any imported image to its own format - typical MS "I'm in charge" attitude.
I avoid .JPG like the plague - every time you re-save a .JPG image you lose some detail and create 'artifacts'. Much better to convert to .PNG - even if your camera/phone initially exports a .JPG file.

AES said:
All of which has nothing to do with your EXCELLENT WIP. It's too late for me and this toy car, but next time ("Next time????? I haven't even finished the current one yet!!!") I shall definitely try the idea of your "planning/diary" approach - that would have saved me from a couple of problems with my "mods with un-for-seen consequences"!!!
I haven't updated my WIP recently because I've been correcting some errors I made early on and it's taken me far too long. I have also been considering whether it is still beneficial since there has been very little feedback so your kind words have 'turned' that.

AES said:
BTW, I might have missed it, but I saw somewhere above that someone in your family had tested Covid positive but didn't see anything else. I do hope you and all yours are OK.
Yes - my grandson (who lives with me) but neither of us have had any symptoms and have completed our self isolation.

AES said:
Thanks again and if it doesn't sound too patronising" keep up the good work.
Your post has had a positive effect so look forward to an update shortly :)
 
Thanks J-G.

If I may, a Q please. Using my old (V5) version of Photoshop Elements, I've just tried to save an originasl (straight off the camera) .jpg photo as .png, as you suggest above. When doing so a small options box appears asking me if I want "interlaced" or "not interlaced". Which should I choose please? I've just looked in my O'Reilly book about PS E V5 and can't find the answer. Can you advise/explain please?
 
Spurred on by AES I have spent the morning trying to recall my activities over the past week :) I see that my last .PDF has not been viewed so that may indicate that Don is not currently able to read the forum, I do hope that does not mean that he is unwell.

An overlapping post from AES has just come in --- The point about 'interlaced' is to do with how quickly an image is displayed when being viewed on the internet. An interlaced image will show every other line of pixels - top to bottom - before returning to the top and 'filling in' the missing lines. On modern PCs and fast internet there is little difference and I often forget to select 'Interlaced' when the image is intended for the internet rather than local display but in general I would select interlaced for any web content.

===========================================================================
9th Dec
Having made eight new 8T Pinions and fitting them to the spindles that were to hand I realized that there were another four 8T Pinions needed as well as four 10T Pinions. These are for the gears which were not in the workshop simply because I needed more space and I’d forgotten them.

This meant that I had to create another drawing, DXF file and G-Code but having done multiple Pinions before I found that I changed the way that I handled the G-Code. Previously I had cut away a circle around the outside of each Pinion and then used a drill MOP to take away the main bulk of the space between the teeth only later using the outline to first do a rough cut and then a finishing cut. This time I used only the outline to first cut a swathe with a 4mm end mill leaving the teeth 1.3mm oversize followed by a 2mm burr leaving 0.4mm and finished with a 1mm end mill. The first two went 7mm deep and the last one 6
Fig-153.png
mm. The Pinion leaves need to finish at 5mm thick so there would be ample ‘meat’ to trim both faces on the lathe in the event that the top face sustained some damage - which I don’t think it has.

This method I think is much more efficient and ultimately produced a block (Fig-153) which will be simple to separate on the small bandsaw.

10th Dec
That still took some time and I needed to be cautious when trimming the excess from individual Pinions! Removing the inner screws from the Ash jaws and boring them out to glue on Maple pads was definitely worth while and made gripping these Pinions for the second operation much better.

The new - slimmer leaved and smaller - Pinions are definitely better but to be absolutely sure that they will run freely I need to glue up the frames so that I can fix the locations of the bearings. The jig (read - flat board with screw blocks!) that I used to glue up the Dials was easily re-worked to apply pressure to each joint but it’s a one at a time
Fig-154.png
job so will take a day or two to complete the 8 frames.

11th Dec
When I spoke about gluing up the Dials, I mentioned that I would use a 50mm Ø sanding disk but forgot to take a photo – Fig-154 shows the same process but used on the frame. The pad is held in the collet on the mill and the height adjusted to suit the distance between the flat board (40mm Formica faced kitchen worktop offcut) held in a vice. I started with 60grit and will eventually get down to 320g.



17th Dec
The past few days have been spent correcting some errors I made due a naïve belief that it would be wise to cut all the recesses for the Bearings etc. at the same time that the frame sections were made. It certainly seemed sensible at the time but once the frames were glued together it became obvious that the exact location of the bearing positions needed to be determined relative to the main spindle (which carries the hands) and with the knowledge of the mesh of the Pinions & Gears. Even a difference of 0.1mm in the position of adjacent spindles can make a difference to a clock running for 5 minutes or 24 hours.

I was very pleased with the way that the frames went together as far as the joints mating was concerned but the position of the winding spindle holes caused some concern. This was primarily to do with the short Section D joint being an ‘easy’ fit. Add to this the issue that necessitated the re-cutting of the Pinions and I decided that it would be prudent to simply bite the bullet and re-cut all the bearing recesses, along with the positions of the frame spacers.

A great deal of head scratching was needed and some experimentation which ultimately detracted from the process of photo recording - ie. not being clear in my own mind what I wanted to do I resorted to my ‘norm’ of just getting on with the job in hand. So, I’ll add some drawings to show how I solved the various problems that arose.

Because the Gear/Pinion mesh would benefit from ‘a little extra’ separation, (which I had determined by trialing the fit of the hour/minute planetary gear pair) I decided that the most efficient way to reposition the bearings would be to machine a recess and insert a new piece of wood into which I could cut a new bearing recess. This would allow me to add 0.2mm to the centre distance. Greater than this and I would be in danger of compromising the escapement. The position of the hour/minute planetary gear doesn’t matter quite so much so I made that 1mm greater than originally designed.

The Denford has a maximum capacity of 400mm in the X Axis and 200mm in the Y Axis so cannot totally accommodate the assembled frame without some judicious positioning. Fig-155 shows how I positioned the frame to cut recesses for the new bearing inserts. I chose to use Maple and make this correction a ‘feature’ :)

The black blocks are positioning aids which were machined with reference to the table centre locating peg (Red Cross). The Frame was kept in position by fitting three clamps.
Fig-155.png

Using this set-up I could machine the three Bearings along with the Hour/Min Axle and M5 Thread Insert 3. To cut the other Thread Insert recesses I had to re-position the frame as in Fig-156 which took a little more ingenuity as far as accurate positioning is concerned. By now the Bearing recesses had been machined and I knew exactly where
Fig-156.png
Bearing 3 had to be so could create another G-Code file to cut a matching location hole in an overhanging Jig Block. At the time of writing this I had already taken a photo of a similar Jig Block so although out of sequence, Fig-157 will give you a better idea how this ‘overhang’ Jig Block allowed me to use a locating ‘Peg’.

D’oh! - preparing this photo I discovered that I had taken a photo of the first positioning Jig Blocks - but for a different purpose - so that will be Fig-155A.
Fig-157.png






My original intention was to screw the M5 Threaded inserts directly into the frame but once I’d decided to use inserted blocks to reposition the Bearings, it became obvious that doing the same with the Inserts made more sense. It is important that Threaded Inserts are screwed into ‘face’ rather than ‘end’ grain and turning small blocks of face grain timber on a lathe has its own limitations so after making a prototype by hand, I created a G-Code file to make 12 from an off-cut of Walnut. I didn’t bother with cutting the outside circle, just machined 12 holes with matching recesses and part cut separating slots. This allowed me to screw the threaded inserts in on the mill so that I could then mount them on the lathe to machine to finished size which would guarantee that the insert was correctly centred when glued into the frame. Fig-158 is a montage of two photo’s taken from top and bottom of the mounted insert after part machining.
Fig-158.png


Writing this ‘after the fact’ has meant that I’ve missed out some salient details as well as not taking photo’s!. Specifically I haven’t mentioned how I made the Maple inserts which had to be glued in place and sanded down to a level surface before machining the final recesses for the bearings.

This was a simple matter of taking a 22mm wide strip of 8mm thick Maple and writing a G-Code file to cut 4 x 20mm Ø and 8 x 12mm Ø ‘buttons’. They were cut 7mm deep and then separated on the small bandsaw. Once the ‘buttons’ were glued in place the frames were returned to the Mill to have the excess sanded off before going back on the Denford to have the real Bearing recesses machined. This was also necessary for the Thread Insert repair blocks of course though they didn’t need a second operation.
Fig-159.png


Having spent all morning trying to remember what I’ve actually done over the past week, I think I am nearly up to date so can at least add a photo of the current state of the front Frame (Fig-159) showing all the corrections.

I now have to do the same with the Rear Frame which will entail solving all the same problems again but in the opposite hand! Hopefully it won’t take as long!

Before posting this update I’ll add a photo of the Front Frame but with the spacing rods and some of the gears/pinions in place.

Fig-160.png
 

Attachments

  • WIP-31.pdf
    3.1 MB · Views: 6
Thanks a lot for the info re .png J-G. Noted n understood.

Re the .pdfs, if donwatson is the same as me, I reckon what we're both doing is "speed reading" your on line stuff (i.e. mainly looking at the pics & glancing at the text n dates) then downloading your .pdf s for later perusal. I presently have I think 3 or 4 of your .pdf s that I haven't even opened yet (getting near to Xmas, the number of "honey do's" seems to grow exponentially - a bit like Covid really!

(If I can make that "joke" at this time without offending anyone - NO offence intended - e.g. here our "R number" is above 1.0 and still increasing and a lot of our hospitals are apparently really close to their staff limits, while out National and Kantonal governments argue with each other about what to do - pure political posturing IMO - so I'm really NOT making the above "joke" lightly).
 
Having now got my own Denford I'm taking a bit more interest in the machining aspect of this project. Up till now I've been more interested in the clock itself, so it's nice to see it coming together.
Brian
 
Thanks for posts JG. I had minor irritations and am now back. AES has hit it on the head when he describes what he does, that is my way also.
Originally I was interested in the clock itself with a side interest in the CNC machining but I have become totally absorbed in this WIP and have revived my interest in using my Stepcraft machine to maybe try another project.
Thanks again for this WIP it has been great working through you problems.
take care and stay safe
Don W
 
Yeah, I've just "discovered" that I've got your WIP nos 28 to 31 on my hard drive and not even opened yet. It appears from looking at your actual pages, the downloads don't count as "read" on the note-thingy against each .pdf, so don't be disheartened Mate. "People" (well at least Don & me anyway!) ARE reading - AND enjoying.

Seasons greetings, and stay safe & well.
 
Back
Top