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.
My workshop time over the weekend was restricted due to F1 & Strictly :) . . . but I did get some work done.
=====================================================================
23rd Oct
The drawings for the 32T gears were done a while ago but this morning they were still in CorelDRAW! I’m pointing that out only because it astounded me that I hadn’t created the G-Code and sent it to the workshop until mid-afternoon. It’s astounding how long it takes (me) to arrange multiple identical items on a grid to allow for minimal tool clearance and avoid clamping holes!
Fig-87.png
Maybe I’m being too ‘tight’ in trying to squeeze the most out of any piece of timber and not considering the time element as part of the cost....... hmmmm.

I’m finding this Maple to have a mind of its own! As you see in Fig-86 the blank was screwed down and flat against the table but, after the outline of the gears had been cut, the ends had lifted up by 2 - 3mm so before I did the finishing cuts I needed to use another 4 clamps to make sure it remained flat. I didn’t need the clamps after turning it over and the trimming of the backside made it easy to release all eight with just a few strokes of my piercing saw.

There were periods of 40 - 50 minutes where the Denford simply got on with following the G-Code paths so I used that time to mill the flats on all the spindles plus drill & tap the Winding Drums.

Finished the day by cleaning the 32T Gears to 5mm thick on the lathe - I still have to machine the recesses to take the spacers but I thought 11:30pm was quite late enough to be in the workshop! I suppose that Fig-88 is not a bad days work though, even if most of it was done previously :)

Fig-88.png
24th Oct
Finally! - I bit the bullet and broke out the Table Saw to cut up the Walnut board. My original plan was to simply cut the 9½” wide board down to just over the 6" that I needed but after looking at the options decided that cutting a 700mm length off, that would be easier to handle than the 2.2m long board. It meant that I wouldn’t be left with a 2.2m length of 3 x 1½ Walnut which might become useful for some future project, but considered that a small price to pay for the convenience of handling a smaller ‘lump’. In a similar fashion I then cut another shorter length, enough to cover the next set of parts and so on.

Although most of the frame components finish up at 10mm thick I’m only taking two pieces out of the 42mm thick stock at 15mm each and to ease the burden on the band-saw I’m starting the thicknessing on the table-saw by cutting 1/3rd of the depth from each side and then finishing the split on the band-saw. The parts for the Dial are more of an issue. They finish at only 5mm thick so I’m hoping to get 4 ‘slices’ out of the 42mm. With a 3mm kerf that could be challenging so I may attempt to try it on the Band-saw but that may not like an 80mm wide cut.

Time ran out on me today since I had other chores to get done but I hope to have all 50 blanks ready for their next operations before Sunday is done.

25th Oct
I thought I’d completed the ‘slice-n-dice’ this morning but when I stacked and labelled what I’d got there were only 48 pieces and I expected 50 ------- it took me some time to work out what was missing and I discovered that I’d missed one item off when I created the cutting list so in reality I needed 56 altogether. Fortunately there is plenty of spare so the extra 8 were soon sorted.
Fig-89.png
The facts that the Band-saw was causing me some grief and I really needed it to play ball when I was cutting the slices for the Dials made me spend some time looking for the problem.

I had fitted a new blade less than a fortnight ago but I did have another on the shelf so that was certainly worth fitting. More important was the fact that I couldn’t move the guide assembly down to just clear the work so that meant a major dismantling of the mechanism. It’s a very questionable design since the only way to get it off the saw is to drop it to the lowest level before removing the top wheel, but since saw-dust can (and does) get into the Rack & Pinion jamming it up it’s impossible to get the wheel off. I had to trim part of the structure away so that it didn’t foul the wheel. Even then it is very tight against the top wheel casing needing a specific orientation before allowing the clamping stud to go through the clearance hole. Drilling that hole out to 11mm rather than 9.5 made all the difference.
With the new blade in place and the guides at a new low, cutting 8mm thick slices for the Dial, whilst not a ‘breeze’, was certainly much easier with little to no ‘wander’. I had been taking it very slowly (specifically with the 150mm wide Maple) but actually found that a firm ‘push’ improved matters.

26th Oct

Naturally, cutting the Walnut board into the basic sawn blanks for the Frame, Dial & Wall Plate is only the start of process and since my Table Saw is not an industrial Wadkin or similar I needed to make sure that the edges of these blanks were square to the faces so that when clamping them in the ‘vice’ on the Denford table they wouldn’t tilt - I’d made a decision that to skim the surface it would be better to clamp them by the edge rather than drill recessed holes so broke out the eccentric clamping block that came with the Denford. This meant that all the blanks now needed to have the edges cleaned up on the Router Table.

It also seemed sensible to fix a piece of scrap to the table and machine the face of that to a specific point for each component and Fig-90 shows both the ‘Stop’ and the clamp.
Fig-90.png
Fig-91.png
As I don’t have a belt sanding thicknesser and most of the blanks are only just over 200mm long - too short for a P/T but too long for my mill - the surface needed to be cleaned up on the Denford. Starting with a blank for one of the side spurs of the Frame I soon realized that it would be most efficient to clean the surface of all of these blanks before moving on to cutting the outline and joints.

Cleaning up 16 blanks with each pass taking only 2 minutes meant that I couldn’t just leave the Denford to get on with it so I only got to cut the first outline and joints on the first blank quite late in the day and I forgot to take photo’s before I’d done two! The prepared blanks are stacked in Fig-91 and the first strut with joints cut in Fig-92. I still have to turn that over to cut a recess and finish the surface to 10mm thick.
Fig-92.png
 

Attachments

  • WIP-17.pdf
    2.1 MB · Views: 3
27th Oct
Today’s been a bit fraught with the Denford trying to make me respect it more :( The four rear right struts are cut out to depth and the first of those has been turned over and machined down to just over the finished 10mm thick. My first attempt at skimming the second face had to be aborted because the clamping mechanism I thought might work wouldn’t hold the part finished item flat so I had to resort to drilling and counterboring holes to improve that - or rather make it possible - by screwing the part finished blank down. Fortunately I found three places where I could align the holes with the 40 x 40 grid of holes in the Denford table but unfortunately when I got close to the final thickness, a part of the blank began vibrating causing some chatter tool-marks to appear. Not so much that they can’t be sanded out but enough to make me look at finding another clamping hole. I can’t show an image since my phone/camera can’t see the damage - I might try using an endoscope tomorrow to see if that will pick it up.

The annoying part is that the next operation - cutting a recess - had to be abandoned. At least I can take comfort in the fact that this is exactly what the project is about - finding out how to handle materials in a CNC environment.

28th Oct

Having slept on the problem of the uncut recess, I think I can solve it by re-designing the G-Code. The issue is that the depth of the recess has to be 3mm from the top surface and although the blanks are all about the same thickness the potential .2-.4 mm difference does matter so that was the reasoning behind trimming to the final thickness first but in fact if I use the table surface as the datum, then I do know that the recess must also be 7mm from there. So, I can set the G-Code to cut a recess stopping at +7mm from the table rather than what I currently have which is -3mm from the top surface - the end result will be the same.

This tactic also obviates the need to use the 40 x 40 grid holes to hold the parts. The part finished one can be positioned just as easily as the other three. Once the recess has been machined on all four I can go back to the surfacing operation using the clamping holes including the one which has been recessed though that won’t be on the grid so will need a wood-screw into the table which won’t matter a jot.

Having cut the recess I took a photo’ to show the clamping strategy but then thought that I could also try-out the endoscope to see if that would pick up the ‘chatter’ damage. It took a while since I had to install first Zoom, to give the Endoscope somewhere to display the image, and then Photoshop so that I could display the screen-grab.

As it happens, the chatter damage can be seen on the Phone Photo’ (Fig-93) but the Endoscope image (Fig-94) makes it more visible.

Fig-93.png
Fig-94.png


The four Rear Frame E Struts are now both recessed and skimmed to just over the 10mm finished thickness and I’ve started on the G-Code for the Rear Frame ‘C’ strut which is the opposite side. Having found that clamping down to the table via the grid of threaded inserts is necessary I’ll include the clamping holes in the G-Code and, bearing in mind the ‘chatter’ problem, modify the [Holding Tabs] from the 3mm x 3mm triangular default to 4mm x 4mm and square which will give a stronger hold for the second side machining. I’ll still have to cut the counter-bores as a second operation on the mill since they have to be from the side that is against the table but the positioning of the screw hole will be accurate and the counter-bore is by definition oversize and not critical.

Although I have the blanks for the Front Frame C & E struts prepared, I think that it will be better to prepare the blanks and finish the Rear Frame D extension first which will mean that I can get the whole Rear Frame put together.
 

Attachments

  • WIP-18.pdf
    622.8 KB · Views: 3
What a morning!!! I did some work first thing and updated the WIP with the intention of posting it at about 10 o'clock. First problem was that my PC wouldn't allow me to create a .PDF file ??? The printer selection kept reverting to the physical printer which is the first in the list but not the [Default].

It seemed that a keyboard key was stuck because, when I tried to remove and re-install the .PDF Printer Driver, the file listing kept moving to the top of the list ??? Eventually traced it --- after closing all the open programs and re-starting the PC - twice!! --- to the wireless mouse/keyboard USB receiver which I removed and re-inserted and that sorted the problem.

I've been building PCs and providing support for over 30 years and never come across such a silly issue before.

I still had some issues with the .PDF losing some text so had to do two separate pages.

2 hours later and I can start to transfer the latest installment!
============================================================================
29th Oct
Having done all the transfer of CorelDRAW! drawings of both sides of the Rear Frame ‘C’ strut to .DXF files I soon had the G-Code files ready and could let the Denford & Mach3 get on with machining them while I got on with other small items such as spacers on the lathe. I had to keep checking the Denford to change tooling of course but I was able to sort of multi-task.

All went well for the first strut but just after one pass cutting the outline of the second I was aware of a change in the sound - even though I don’t make a point of listening to one machine doing its own thing whilst I’m working on another, the natural instinct is to be ‘aware’. The change came about because the 3mm cutter - a cheap single flute TCT item - had broken off just below the shank :( ie. no cutting edge at all. This was a 1mm deep cut in wood so no real stress. Not having another 3mm router bit I had to re-jig the G-Code to use a 4mm.

The next problem arose when I left the Denford machining the last blank while I had a bite to eat. You can imagine my dismay when I got back in the workshop to see that the blank had moved about in the X axis due to being just a little narrower than the other three and not being held as firmly. ( I became complacent!!) A piece of thin card soon sorted the clamping out but I had to guess at the precise lateral position and although I did complete the remaining oper
Fig-95.png
ations I won’t know if the strut can really be rescued until I come to fit it to the rest of the frame. Fig-95 shows the damaged strut at the bottom and compare it with a good strut at the top. If I can make it fit then I’ll still have to ‘bodge’ an in-fill to repair the damage. The lesson learned is that I must also clamp in the X axis.

Other than that, I have the spacers made for the 32T to 60T gears (though not the 60T Gear yet) on the spindle which will also carry the hands. The hole in the spacer needs to be a good fit on the 6mm spindle to maintain concentricity so it was drilled out to 5.5mm and then Reamed to 6mm.

Because these spacers are
Fig-96.png
made from Maple, I anticipated that even reaming them might leave them ‘tight’ due to wood fibers swelling so I put the reamer in the lathe head-stock so that I could re-ream from both sides - working the spacers by hand until I got a good press fit on the spindles - that way the wood fibers were cut from both directions thus minimizing any tear-out.

I still have to ‘finish’ the teeth on the gears but that is going to be an intensive ‘hand-work’ job for which I think I’ll make a ‘frazing stick’ with 320g abrasive.
30th Oct
This morning I’ve completed the second side of the ‘C’ struts which gave me little time to work on the smaller parts which have to be done on the lathe but I did get the motion works spacer started before returning to the ‘D’ section of the Frame, specifically to machining the blanks to an even thickness. I’ve now realized - because I started to create the G-Code - that I’ve prepared the Front ‘D’ frame rather than the Rear. No big deal since both have to be done anyway but I was working on the Rear Frame, so I’ll create the G-Code for both before getting back in the workshop to prepare the Rear ‘D’ blanks.
Fig-97.png

Not a brilliant photo’ (Fig-97) but you c an see that the Frame ‘C’ strut has held together due to the wider & higher Holding Tabs and that the Rear ‘D’ blanks are taller than the Front ‘D’ blanks. The only difference between them (at this stage) is the width. The Front one has to take a larger bearing so is 42mm wide whereas the Rear is only 32mm. Ultimately the difference is also that the Front D Frame can have the joint and the bearing recess machined from the same side but the Rear D has to be turned over to cut the recess so I need to take care when selecting the position on the table. That difference made me realize my error.

A number of issues about work-holding came to light when creating the G-Code and it would probably have been better to leave the Walnut for the ‘D’ section of the Frame in one length rather than cut to individual blanks. Hindsight is a wonderful thing :( Tomorrow I’ll have to see if I can machine the Frame D parts with just clamping in the Y axis as opposed to screwing them down.

31st Oct
The Rear Frame D sections are now all ‘thicknessed’ using just the eccentric clamping jaw. They cleaned up at just under 13mm (the Front Frame version is just over 14mm) so the G-Code has been adjusted so that I don’t cut into the table.

I’ve been cautious about using a 2mm burr to cut the outline due to the 11mm depth but since the 3mm TCT cutter broke leaving me only a 4mm option which I considered too wide for this part, I decided to specify it when creating the G-Code but I reduced the depth of cut to ½mm rather than 1mm which I’ve been using for the ‘rough’
Fig-98.png
outlines with the 3 or 4mm TCT cutters. I also specified the 2mm burr for the finish cut since my 1mm spiral flute cutters can only reach a depth of 6mm. The only problem with using burrs instead of spiral flute cutters is that they leave ridges. These will be removed when I do the finishing process through the abrasive grits of course.

As you can see from Fig-98, I’ve re-thought the clamping methodology and made specific width spacers to achieve a good grip but also restricting the blank in the X axis as well. Using this method also affords me the opportunity to cut the joint on all four blanks before changing the tool to cut the outline or drill the hole. Previously I’ve kept the blank in place and changed the tool for each operation before starting again with the next blank. For this particular part that would have meant 12 tool changes rather than 3, so overall quite a bit more convenient and faster - not that I’m too concerned about speed of manufacture.

Machining the outline was taking 15 minutes so there was time to do other work on the lathe
Fig-99.png
and I got all the Beech bushes for the motion works made though not the brass bushes which will provide a ‘free running’ bearing for the hour hand.

Because I had no means by which I could screw the ‘D’ blanks down to skim the second side, I had to use a clamp at the right hand end. I was well aware that I couldn’t run the fly-cutter straight across the blank so when I created the G-Code for the surfacing I made sure that it stopped short of where the clamp would be. That would be all well and good as long as I also made sure that I positioned the blank in the correct position on the table. The adage of “measure twice and cut once” comes to mind here!!

Fig-100.png
What I actually did with the first blank was position it 40mm (one grid hole) too far along the X axis. I should have done a ‘dry run’ by cutting air but unfortunately just watched as the Fly-Cutter ploughed straight into the clamp and broke both carbide bits :( Fig-100 is the remains of the carbide cutters retrieved from the cutter block. See WIP-3 Fig-12 for the cutter block itself. As it happens I use - and break - quite a few 1/8" Ø solid carbide drills so I have a stash of spare ‘shanks’ so it only took me 10 minutes to replace the bits and sharpen them.

1st Nov
I’m somewhat surprised this morning to find that the blank is in the correct location on the table ????

This means that I had to look again at the CamBam parameters/drawing and it turns out that I had drawn the engrave MOP line just short of the furthest extent I wanted to cut to rather than to the radius of the cutter - D’oh!!
Fig-101.png


In Fig-101 you can see the damage the fly-cutter did to the clamp and the Rear Frame D machined to 10mm thick as far as the cutter needed to go.
It also shows the ‘holding tabs’ which keep the integrity of the clamping sufficiently to take surface skimming cuts - I tend to keep them under 1mm deep.
 

Attachments

  • WIP-19A.pdf
    839.2 KB · Views: 3
  • WIP-19B.pdf
    1.2 MB · Views: 3
EXCELLENT WIP J-G, thanks. I'm really enjoying this, thank you very much Sir! 👍

While definitely NOT wishing to gloat ('onest Guv!) being a definite PC klutz myself I was heartened to read that the experts have "funny PC problems" themselves from time to time.

Keep up the good work please - I know myself how much time and effort goes into producing such posts.

All the best.
 
Thanks AES - it's difficult to know how useful - or simply 'interesting' - a WIP is when there is little feed-back so yours is much appreciated.

Being able to solve PC problems is often down to having experienced the same (or very similar) problems yourself and most of my customers know me because I've suffered the same issues - and sorted them - in the past. It's very much a matter of following the 'logic', - after-all PC aren't sentient no matter how often it seems that they have their own agenda :)
 
Being able to solve PC problems is often down to having experienced the same (or very similar) problems yourself and most of my customers know me because I've suffered the same issues - and sorted them - in the past. It's very much a matter of following the 'logic', - after-all PC aren't sentient no matter how often it seems that they have their own agenda :)


Thanks J-G, glad to offer "encouragement" :cool:

Re your comment about PC's (quoted above), sorry but you're quite wrong you know - ALL my PCs most definitely DO have their own agenda (or perhaps better said, "that young Mr. Gates and his merry band of brigands do"!!!!!!!!!!! :mad:
 
Starting to see some progress now that I have a Frame 'Dry assembled' so here is the 20th installment!
==============================================================================

Once I’d finished the Rear Frame D struts it made sense (to me anyway) to work on the FRONT Frame D struts since they were already prepared by having the surfaces skimmed and they would be easier than those for the Rear Frame as the recess for the bearing is on the same side as the joint so there was no issue with re-positioning after turning over. It still took me the rest of the day to get just the one sid
Fig-102.png
e done due to other commitments! — and even then I’m not satisfied with the result. You’ll see in Fig-102 that I mis-judged the position of the blank on the Y axis by about 1mm. Fortunately the diameter of that section is not ‘critical’, 40mm was just a convenient number I came up with so I can quite easily create another MOP to cut it down to 38mm.

The only other part I’ve done any work on today is the 10T Pinions used in the Motion Works - the start of the train that takes the hour hand rotation and speeds it up to drive the minute hand - I thought I’d done all the work on these but I’d been less than explicit when I did the technical drawings and I mis-read a 12mm x 1mm [Slot] as a 12mm Ø [Recess] so that needed to be addressed. These 10T Pinions fit on the main spindle but in front of the frame and pass through a needle roller bearing. They are drive
Fig-103.png
n by a 1mm Ø pin which fits in a cross-drilled hole in the spindle – which I can’t drill until such time as I have both Front & Back Frames fitted together – and that pin sits in the slot. The recess I cut in error won’t matter because it will be behind a larger gear.

In Fig-103 you can see the slot in the stacked pinions and the recess in the pinion on the spindle. There is also a slot in the single pinion but it’s difficult to see.

2nd Nov
Doing the 38mm Ø trim on the Front Frame D strut has taught me that I don’t need to do both roughing and finishing cuts with the 2mm burr. You’ll see in Fig-104 that the finish straight off the burr is very good indeed - though trying to photograph the ‘ribbed’ finish is not easy -
Fig-104.png
this is on the straight sections which were cut with a ‘finishing’ cut of only 0.3mm but the main reason for the ribbing is down to taking a 4mm depth of cut. This would be fine if I were using a spiral flute cutter but I suspect that the ‘burr’ has a 1mm pitch and the ‘teeth’ are not cut on a helix so it is effectively a ribbed cutter. The trim to 38mm Ø on the round section was done with a 0.5mm depth of cut which I presume gets around the burr ‘pitch’.


Now all the rear frame sections are cut out I spent the morning breaking away the waste, removing the holding tab left-overs and dry assembling the parts to check that there are no major adjustments needed. I’m well pleased with what I found. The C & E struts all went together with just a little hand pressure. The D strut is an ‘easy ‘ fit and that is - if anything - a benefit since it will allow me some tolerance in the positioning of the winding spindle bearings which can only be finally fathomed when I have the two frames fully assembled and the main spindle in place. I’ve annotated Fig-105 to give you a better understanding of how I referenced the struts. I
Fig-105.png
’ve also shown the frame with the brace made from Maple which again fits very snugly in place - ultimately it will be glued both to the rear frame and also to the spacing pillar at its foot which will hold the front & rear frames apart.

I did find one small error though. For some peculiar reason when I made the ‘A’ component I cut the main spindle bearing recess 8mm deep rather than 3mm ??? - 8mm is correct for the FRONT frame which has an 8mm long Needle Roller bearing but the Rear Frame has a 3mm thick deep groove ball bearing.

This meant that I had to make some 8mm Ø x 5mm long plugs which I did from some of the Walnut scraps.
 

Attachments

  • WIP-20.pdf
    1.7 MB · Views: 3
3rd Nov
A busy and interesting day with a few mistakes made & corrected. It started when I picked up the sawn blanks for the Front Frame ‘A’ section and began to think “Why did I need to saw these at 15mm thick when they will finish at only 10mm?”. This question was answered some time later when I realized that I’d created G-Code to cut both joints 5mm deep when in fact one of them should only be 3mm deep. Not only had I created the G-Code, I’d actually cut all four before I realised :(

This meant that I had to trim 2mm off the face and then re-cut the one joint back down to 5mm. Re-positioning wasn’t a problem with that last one which I hadn’t dis-mounted but the other three did need careful attention to make sure that the width of the 5mm deep joint wasn’t increased.

The fact that I’d (unusually for me) allowed a 50% waste factor saved the day but it took me all day to recover the error and I eventually completed the four Front Frame ‘A’ sections just before midnight — I had taken time out to eat though :)

I was hoping to get the frame spacers turned while the Denford was busy and I did at least get some Maple sawn to 18mm square and used the Router Table to round the corners so that I could hold the blanks in a 20mm collet chuck. Turning them down to 16mm Ø and drilling a ¼” Ø hole will be a job to do the next time the Denford is tied up.

4th Nov
I hadn’t finished the Front Frame ‘A’ section yesterday, because the cut to round the ends only went 1mm over depth so I still had to remove the waste and clean them up on the linisher.

For some time I’ve been concerned that I didn’t do a thorough job when I drilled the grid on the Denford Table since I only drilled the holes deep enough to take the threaded inserts and I’ve found that the M4 screws which I have in 5mm increments sometimes ‘bottom out’ before the work had clamped. I had been manually drilling the table deeper for the holes that I needed to use and then tapping them, but this morning I resurrected the G-Code which cut the grid and changed it so that all the holes were drilled right through the sacrificial table and left that running while I created the G-Code for the Front Frame ‘B’ section which will be the next part to be made. I’ll still have to tap each hole by hand of course but I can do that on an ‘as-I-need’ basis. At least I can’t make the same mistake as yesterday since all the joints on the ‘B’ section really are 5mm deep!

OUCH!!!

Well, I didn’t make this mistake! It seems that either Mach3 or the Denford decided that I was doing too well today having skimmed both surfaces of the Front Frame ‘B’ sections with nary a problem, and even cut the joints and two holes in the first blank whilst also giving my grandson instruction in the use of the milling machine during which time I just let the Denford get on with cutting the outline. I knew it was going to take a while since I’d decided to take just a single cut down to the 11mm depth in ½mm increments and I just kept an eye on what it was doing occasionally. All was going to plan and the depth was at about 8mm when I had to spend a little more time on the mill. The next time I looked at the Denford it was cutting the blank in half !!! - The drawing in CamBam, Fig-106, from which the G-Code is generated has no line going from top to bottom but that is what was being cut when I got back to checking the Denford – the result can be seen in Fig-107.
Fig-106.png


Maybe the Denford felt neglected :)
Fig-107.png


I’ve poured over the G-Code and can find no error so therefore no reason that the Denford suddenly changed from following the instructions and simply moved the table along the Y axis, and since I wasn’t watching at the time I doubt that I’ll fathom what went wrong.

I stopped it before it had completely cut deep enough to sever the two halves and I suspect that I can glue in a 2mm thick piece of Walnut to recover it. The concern is that it will happen again on the next blank so I’ll have to watch every pass - just in case. It will be prudent to split the cutting of the outline into two sections, the first going down to (say) 8mm in 1mm increments and the second starting at 8.5mm, going to 11mm in ½mm increments. That will at least address the issue of the ribbed finish that the burr creates when using full mm pitch cuts and might stop a repeat ‘rogue’ cut.

After some deliberation - read :- ignore the problem and do something else - I switched the Denford off and closed Mach3, left it for a while then re-started both. The first thing that has to be done on starting Mach3 is to [Zero] the axes. This showed that the Axis registration was way out and I now suspect that the ‘X’ Axis motor or the ball-screw had jammed but the ‘Y’ Axis continued to operate.

I have plenty of thin sections of Walnut so it didn’t take long to trim a piece down to 2mm and glue it in to the damage and that is now curing overnight.
 

Attachments

  • WIP-21.pdf
    1.6 MB · Views: 1
Last edited:
5th Nov
Re-created the G-Code for the outline in two parts and the first run of the 0-7mm deep code went exactly as expected. The depth of cut at 1mm was no problem and because the Y axis only cut must have been 8 or 9mm deep without breaking the 2mm burr I suppose that I could quite easily use a 4mm cut but I’m not yet confident enough to attempt that!

The main cause for concern is that I have no idea how the X axis lock-up happened so can’t tell if or when it might happen again. This means that I have to keep an eye on it rather than leave it working away while I do other work. Only time will tell and I’m sure that I’ll again become complacent and just get on with lathe work.

Not quite ‘complacent’ but I have been making the Frame Spacing struts which I started a couple of days ago whilst still watching the Denford. Rather than drive them with a Collet Chuck (they wouldn’t fit in the largest I have) I turned them ‘between centres’. A total of 12 ar
Fig-108.png
e needed (3 per clock) - 16mm Ø and 78mm long with a ¼” hole. There needs to be some small clearance to allow for minor adjustments and the clamping screws will be 6mm Ø with M5 threads on each end. I made 4 lengths at just over 300mm long so I could afford to grip the first 20mm in a carrier (Fig-108) and still cover the 3 - 80mm lengths.

The Front Frame B sections are mostly finished but I still have to create the recesses for the Dial Pads and for that I’ll have to make some positioning blocks so that I know exactly where the Dial Pegs should be.

6th Nov
Serendipity reigns it seems because when I designed the Dial, I put the stand-off/location pads at 160mm apart which is a multiple of the 40mm grid spacing on the Denford Table so I can use the existing screw holes and just make locating pegs to fit the already drilled holes in the Front Frame B section. More than that, the head of a standard M4 Button Head screw is 7.3mm Ø and the holes I want to locate are 6.8mm Ø so could easily be turned down to fit the holes. Using CorelDRAW! it is a simple matter to draw two circles at 160mm centres, align them with the grid pattern and export a .DXF file ready for CamBam to create the G-Code to cut just the recesses. So no need for positioning blocks, I just had to put the B sections on the two machined screws and clamp them down.
Cuttin
Fig-109.png
g two recesses only took about a minute so there was no time to multi-task and finish the spacers but I did want that job completed so before starting the Front Frame C & E sections I spent some time on the lathe and completed the spacers. I‘d machined the one end and drilled through as a first operation but I had to trim to length by machining the other end and it seemed sensible to drill out this end to 7mm Ø to make sure that the 6mm shafts would have no prospect of ‘binding’ and even give me some room for minor adjustments in case that became necessary.

In Fig-109 you can see both sides of the Front Frame B section before I separate the component from the blank by sawing through the holding tabs. In the background are the Frame Spacers. The change from using the (CamBam) default holding tab dimension of 3 x 3mm Triangular to 4 x 4mm Square has made second operation surface skimming much less demanding

7th Nov
I thought the
Fig-110.png
Front Frame C section were going to be very straightforward since all the joints are on the same side so there isn’t a potential ‘alignment’ problem but in any case I’d decided that screwing a piece of scrap material to the table and cut it by CNC to exactly where the blank needed to be (Fig-110) gave me a very solid ‘Location’ clamp. I’d also written the joint depth on the drawing in CamBam to remind me that they aren’t all the same - however I hadn’t reckoned with the Denford wanting to assert it’s authority!

The X axis motion failed twice out of the four, causing some damage to one joint shoulder - the same joint in both cases. This may mean that I have to fit a small wedge of Walnut which will unfortunately be at the front but may be masked by the Dial - until I do the trial assembly I won’t know. In Fig-111 the lower component is good and you can see the damage in the top one. I suspect that there will be no more than 1mm taken off the shoulder which should be at a small angle (Green Line) but due to the X axis lock up has been cut square.

Fig-111.png

Naturally I have been investigating what might be causing this problem and have discovered that rather than the X axis motor or Ball Screw seizing up - which would be a serious issue needing difficult intervention - it seems that the X axis Stepper Motor Driver is switched off. I only found this out when I opened the back of the Denford to look at the ‘Electronic’ bits and in Fig-112 you can see that the middle SMD has no ‘light’ - therefore switched off.

Turning the Denford Off and back On brings it back so it seems that it may be some sort of ‘overload’ protection might be being triggered but that is only a guess suggested by the previous owner.
Fig-112.png
I’m now looking at the precise point at which the Driver was turned off and suspect that it might be connected with the location of the [Holding Tabs]. Though why it doesn’t happen on every occasion is weird – inconsistency always defies logic and I may have to raise a question on the CamBam forum.

Due to these X Axis problems I can’t yet comfortably leave the Denford to get on with its job while I do other work on the Lathe or Mill so I’ve not done a lot of work toward making the clamping studs. I have done a little, but cutting an 8mm long M5 thread does take some time so only three have been done, though I have cut them all to the rough length.

8th Nov
Nothing whatever done today - it was spent creating a round for a ‘Zoom’ quiz session that I will be running on Tuesday evening.
 

Attachments

  • WIP-22.pdf
    1.6 MB · Views: 2
I was beginning to think that you'd gone to sleep Don :) the .PDF shows 0 views and I was nearly ready to post another installment at 2am but hadn't taken a photo so that will come within the hour.
 
9th Nov
Back in the workshop, but late so only got 3 of the 4 Front Frame E sections made and found that a 2mm solid carbide burr can be crushed to breaking point by the power of a stepper motor. Good job it broke really, otherwise it may well have punched a hole in the Walnut blank :) As it happens I had two spares so I was soon up and running again. You can see the relative size of the two cutters in Fig-113.
Fig-113.png


I’m still trying to understand what I can get away with as far as feeds & speeds are concerned and for this operation (cutting the outline) I’d increased the depth of cut by 50% to 1.5 mm – to some extent this was prompted, not by wanting to increase work throughput, but to get a clean face with the burr and avoid the ‘ribbed’ face created when using a 1mm DoC. It is removed with a ½mm DoC but that doubles the time taken. The next operation needed the 6mm TCT cutter and the G-Code for that was created with a 1mm DoC and it suddenly occurred to me that if the 2mm burr could handle 1.5mm DoC then the vastly more robust 6mm tool ought to handle 2 or even 3mm DoC. It’s a pity that I didn’t realize this earlier.

On further consideration - can you tell I’m ‘thinking on my feet’? - the DoC that can be taken is not just dependent upon the strength/robustness of the cutter but also the ability of the work-piece and clamping mechanism to withstand the forces........

10th Nov
The second side of the E section is now complete - though I haven’t loose assembled them - so I’ll move on to the Wall Plates. No ‘Joints’ or curves involved but they do have to have pockets and recesses for the Brass ‘Key-hole’ plates, and just for the appearance they need to be chamfered all round - which could very easily be done on the Router Table but the challenge of doing it via CNC is interesting.

I said that there were no ‘curv
Fig-114.png
es’ on this part and that is true of my original design but when I got the first blank on the Denford I realized that it would be very easy to round the corners off and also have a pleasing chamfer following that curve as in Fig-114. To do that by hand or even on the Router Table would have been much more difficult to get right.

Yesterday I remarked about increasing the DoC with larger cutters and the Wall Plate was cut mostly with 4 and 6mm Ø tools with a 2mm DoC - easily handled - the chamfer was also cut at 1000mm/m feed. That compares to the 400mm/m I’ve generally been using.

11th Nov
It was my intention to do some lathe work today but realized that I had to skim the other side of the Wall Plates first and then got to thinking that the chamfering had been easy, so I looked at the Wall Pad which is just a 20mm Ø x 4mm thick ‘button’ of Walnut which will act as a wall protector for the vertical adjustment screw which also has a chamfer. At that size and needing to be ‘face grain’ rather than end grain it is an awkward piece to hold in the lathe but all four could easily be made from a Walnut scrap/offcut - and I have plenty of that!
Fig-115.png

I did finish the 8 Frame Clamps on the lathe while the Wall Plates were surfaced and then set about the G-Code for the Wall Pads and the Wall Spacers - these hold the Steel Buttons which fit into the Brass Key-hole Plates and would similarly be awkward to mount on the lathe - and it is much more material efficient to make by CNC.

Whilst the outline of the Wall Pads were being cut I started on the Brass Cups which fit in the two fixing holes in the Wall Plate (Fig-114) I only got 6 made so the other two will be done tomorrow, along with the Spacers and Steel Buttons.
 

Attachments

  • WIP-23.pdf
    972 KB · Views: 2
12th Nov
I seem to have done a great deal today - but little ‘CNC’. It’s interesting to realize (though I always knew it to be true) that selecting the best method of manufacture is not always obvious. Before I had the Denford if I needed a round wooden object with the grain running across I would automatically cut a square piece - maybe cut the corners off to make it octagonal - and mount it in a four jaw chuck. Having now made the Wall Pads and Wall Mounting Spacers from Walnut off-cuts by drawing them in a ‘staggered’ layout, it makes more sense to do this type of part by CNC - even for a one-off - (though naturally multiple identical parts benefit more) due to the ease of work-holding.

Having finished the last two Brass Screw Cups first operation, I re-mounted each and finished them to thickness and loosely fitted them to the Wall Plates. As I had the 10mm Ø Brass bar to hand, I then made the Vertical Adjusting Nuts. Although they are ‘nuts’ they are 10mm Ø but will be fixed into Walnut spacers with Epoxy. This is better than putting a thread in Walnut end grain even though there is unlikely to be much ‘Adjustment’ when in use - it is just there to take account of the wall (which the Clock is hanging on) not being truly vertical. The Spacers which these fit into came out of 25mm square stock held in a four jaw chuck with a live centre support.
Fig-116.png

In the meantime, the Denford had finished the Wall Mounting Spacers - well as far as could easily be done using CNC - so they could be mounted in a pre-bored chuck on the lathe where they were drilled and counter-bored to suit an M4 Threaded Insert. This was the reason that the grain needed to be across the part rather than along its length - you should never screw into end-grain.

The last part I made a start on today was the ‘Key-hole’ Plates from 1.5mm Brass plate. I did think about making them on the Denford but decided that the Mill was a better option since they are just oblong flat parts with a ‘Key Hole’ and two countersunk holes for screws and the Mill is more robust and can take a 1.5mm cut with a 5mm end-mill in one pass whereas I would have to let the Denford take 8 - 0.2mm cuts.

13th Nov
Fig-116 will give you a better idea how each of the parts I’ve been working over the past few days fit together. They are all concerned with fixing the clock to a wall. The Wall Plate screws to the wall via the two holes with the Brass Screw Cups to provide two ‘hooks’ (via the ‘Key-hole Plates’) for the Mounting Buttons (not yet made) to lock into. The Wall Mounting Spacers will be glued into recesses in the back of the Rear Frame ‘B’ section and the Mounting Buttons screw into the M4 threaded inserts.

The Vertical Adjusting Screw, Spacer, Nut & Wall Pad go at the bottom
Fig-117.png
of the Rear Frame ‘A’ Section so, along with the two Mounting Buttons, provide a three point fixing.

The first operation of the 8 Mounting Buttons is now done (left hand image of Fig-117) so tomorrow I should be able to turn them around and machine the M4 thread.

14th Nov
Well that didn’t go well, or rather I struggled with cutting the M4 threads - not with machining down to 4mm Ø but die-ing the thread. Had I been
Fig-118.png
using En1a material rather than 304 Stainless I’m sure it would have been a breeze but only being able to hold on a 2mm long x 8mm Ø, the stresses involved in cutting even an M4 thread were ‘challenging’ and the part was slipping in the chuck. The fact that I wasn’t feeling 100% health-wise, the workshop was quite chilly and there was both F1 and ‘Strictly’ to watch during the day, meant that I gave up after two had taken me the best part of an hour.

The right hand image in Fig-117 is the finished ‘Button’ and Fig-118 shows it attached to the Wall Mounting Block.

15th Nov
Feeling somewhat better, I tried again this morning but wasn’t happy with the first M4 thread which I could only die about 5mm of the 8 but it did fit in the mating part so can be used. I should have understood from past experience that having only a small section to hold on to would cause problems - with Stainless Steel. Free-cutting En1a would not be such an issue.

The upshot was that I bit the bullet and decided to re-make the Mounting Buttons and cut the M4 thread first. The first two off the lathe was relatively easy and the prospect of holding on the thread - even screwing it into a ‘nut’ - would make second op. machining of the Button face a doddle.

Machinery/components always want to assert their authority though and the third thread I cut made sure I knew who was in charge when I was a little too keen to make sure the thread was fully up to the shoulder — there really is no need for that to be the case since the threaded insert has a 3mm deep clearance — but the last 1/3 turn of the work sheared the thread off - naturally leaving the thread in the die :( Drilling that out was going to be ‘fun’, so I anticipated all the potential problems and started with a 2.2mm Ø drill which did cut about 6mm deep before I turned the die over to work from the other side – that’s when the drill broke !!

I finally did get the stainless steel out of the die by using a short 3.2mm Ø drill and a variety of pin punches to deform the remains of the thread. Can’t show you how or what the problem was because in my ire I simply got the thread out rather than taking photo’s of the mess. The next 5 all went without incident and I’ve now machined the front faces as well.

16th Nov
Today I returned to making gears - well more of the 72 tooth main drive gear - making small modifications to the G-Code in the light of further experience as well as taking account of the fact that I no longer have a 3mm TCT router bit. Knowing that the 2mm Ø Burr will easily stand up to a 1.5mm cut, I didn’t bother doing two roughing cuts for the tooth outline, I did one cut leaving 0.3mm to be removed by a 0.8mm Ø burr. Looking at the first one I made (a month ago!) I decided to increase the speed to 20k rpm (was 15k) and reduce the feed to 80mm/m (from 120). This meant that the final cut of the teeth took 45 minutes but I’m not concerned about time - there plenty of other work to do :)
Fig-119.png

It’s suddenly occurred to me that you may not be familiar with the type of ‘Burrs’ or ‘Up-Cut’ end mills that I’m using so in Fig-119 I’ve put 4 of these alongside a 5 pence piece (18mm Ø) and a ruler marked in mm - essentially to give you a sense of scale. From Left to Right they are 2mm Burr, 0.8mm Burr, 1/8" up-cut End Mill and 1mm up-cut End Mill. They all have a 1/8" Ø shank.

You can see that the flutes on the 1mm End Mill are quite short - in fact it can only cut 6 - possibly 7mm - deep, that is why I’m using the 0.8mm Burr to finish off the gear teeth because the 72T Main Drive Gear is 8mm finished so needs to be cut 9mm deep. I would prefer to use End Mills because they are stronger and will do so for the smaller gears which have still to be cut.

Earlier today I said that I’d increased the speed to 20k from 15k but after finishing the first gear I noticed a little ‘burning’ on some of the teeth so I changed that again to 18k - I also fitted a brand new 0.8mm Burr since the one in use was showing signs of heating. I had occasion to go past the Denford after about 7 minutes into the run and noticed that the new burr had broken!! After fitting yet another, I reduced the depth of cut by 50% as well. That still didn’t stop the new burr breaking so I re-jigged the G-Code to use the 1mm End Mill for the first 7mm and finally the 0.8mm burr for the last 2mm but in 1mm increments.
 

Attachments

  • WIP-24.pdf
    928.6 KB · Views: 2

Latest posts

Back
Top