CNC Clock Build W.I.P.

[J-G]

I had thought about Flashing but didn't look far enough as I thought that H & S had virtually wiped out the use of Lead for flashing due to the potential contamination of rainwater.

A quick Google finds Calder Industrial have a 100mm x 6m roll 1.8mm thick at £33.70 + carriage and that would weigh 12.25kg so a similar cost to the Sash Weight I bought, and it could well be more convenient since the weight could be adjusted by controlling the number of 'turns' used. They also do 150, 180 & 200mm wide.

 

[Yojevol]

My neighbour had a wide gully between two roof pitches relined recently. I was amazed to see them lifting a roll, which must have been 800mm or so, up the ladder!

 

[J-G]

The Lead did not arrive on Friday and FedEx are showing 'Delivery Schedule: Pending' ??? so I'm hoping that it will appear sometime on Monday.

In the meantime I have now made three of the 'Weight Shells', well the tube, caps and

hangers are made but I’ve only made one of the bolts.

I’ve held off making the other bolts simply because as yet I’m not convinced that a 2.2kg weight will be sufficient and the length of the bolt (and the tube) may need to be increased. I say this because the clock has been erratic over the past few days, seldom getting past 5 hours run-time but yesterday I increased the drive weight to 3kg and have - up to now - had close to 8 hours and counting.

I assume from this that I will certainly need more than 2.2kg but maybe not as much as 3kg so I’m thinking that I ought to at least allow for some extra. If there is some space above the Lead filling it won’t matter but at the current length of the tube & bolt I can only get about 2.35kg. Adding 15mm to the current design will increase that to 2.68kg.

On a separate matter, a strange thing is happening to the Pendulum Bob. I noticed a while ago that the orientation of the Bob was ‘rotated’ but initially thought that I must have accidentally knocked it out of position so I then made sure that I checked that before I started the next test. Three hours later I saw that the Bob had rotated about 60°. Whether this is due to some natural phenomenon or the fact that I hadn’t fully tightened the Pendulum Rod to the adjusting screw I’ll have to fathom during more tests.

I haven't done a .PDF of this, I'll add it to the next post.

 

[Yojevol]

Three hours later I saw that the Bob had rotated about 60°

I'm not familiar with the detailed design but could the bob be rotating itself down the adjustment screw thread under gravity?
Brian

 

[J-G]

Hi Brian, I do keep forgetting that I haven't shared the detailed working drawings - there are rather a lot of them

Here's a drawing of the Bob along with how it is attached to the Pendulum Rod.

The 'Front' view shows the Rod, Connector & Screw expanded so you can see that they are separate parts and the 'Side' View shows them in their final position. I've also drawn in a possible solution in the shape of a locking screw which would be an M4 grub screw at the back of the Bob - I'm somewhat cautious about taking this approach since the fact that the Bob is 'locked' to the Rod could be forgotten if ever there came a time to make adjustments.

There is no possibility that the Rod could 'rotate' because it is pinned to the Latch/Finger/Pivot mechanism at the top so a grub screw against the Rod in this way would fix the orientation permanently.

I doubt that the Bob is winding itself down the Adjusting Screw but it is possible that the Connector is being unscrewed from the top of the Adjusting screw. A dab of LockTite might well stop that.

My first thought was "could this be the Coriolis Effect?" which is the reason that I use two pivots rather than one but the time-scale seems wrong - and in fact the twisting has never gone beyond about 60°.

 

[Yojevol]

I would say it is unscrewing on either the connector thread or the adjustment thread and then the walnut binds up stopping it going beyond the 60°. Coriolis effect is extremely small and would be difficult to detect here.
Brian
PS have you checked whether the nut is rotating as well?

 

[J-G]

I did say my 'first thought' was Coriolis - but it was sensibly dismissed

No I haven't yet marked the nut to detect movement so that's a good 'heads-up' and I'd agree that the restriction to 60° could well be a 'binding' - the screw is M6 and the hole is ¼" Ø so there isn't a lot of clearance.

I now have the Lead so can swap the temporary water bottle weight for the real thing - which has a total weight of 2.56kg inc. Caps, Tube, Screw & Cord Anchor - and do some more tests.

 

[J-G]

25th Feb
I’ve already reported that the Lead arrived on Monday and I have a total weight of 2.56kg which over the past few days has kept the clock running - with 3 re-winds - for nearly 2 days. I’m sure it would have continued for as long as I re-wound it but it seemed sensible to move on by shortening the Pendulum Rod so that I could do better tests/adjustments to the Pendulum length to improve the accuracy.

There is potential for 50mm of adjustment within the Adjusting Screw so I first took 60mm off the Carbon Fibre Tube. You’ll see that the Brass ‘Connector’ was epoxied into the Rod; this was advised by the supplier since they considered that it would be unlikely that a thread would remain intact. In fact this is the reason that I used a ‘Connector’ rather than screw the Adjusting Screw directly into the Rod.

Well - since I had a short length of CF Tube to use for a test - I did actually tap it 4BA. This was selected because the bore is 3mmØ and 4BA is 3.6mm OD and 2.81mm ID. It turns out that the Carbon Fibre does stand up to being tapped so I’ve made new Connectors with 4BA threads.

I’ve found a means to ‘Lock’ the Pendulum Bob to the Rod that isn’t as intrusive as a hidden grub-screw. All it needs is a friction device and with the Carbon Fibre Tube is 5mmØ and the hole in the Bob is ¼” there is a 1.35mm gap into which I can insert a ‘Wedge’ (Fig-222) which I’ve made from sliver of Maple. This has the advantage that it will simply loosen if the adjustment is longer (slowing) but would make the adjustment feel ‘stiff’ if it needed to be shorter (speeding) but wouldn’t prohibit any change in the way that a grub-screw clamping that I suggested in my post on Monday would.

Rather than modify the Pendulum that had been in use, I modified one of the other three – this taught me a number of lessons – First, I couldn’t get the clock going for more than a minute. The relationship between the ‘Latch’ - ‘Finger’ - ‘Pivot’ and ‘Escape Wheel Pins’ is extraordinarily delicate and despite ‘fiddling’ with the tips of both the Latch & the Finger, nothing seemed to improve matters.

Eventually I decided to modify the Pendulum that had worked for two days and immediately had it back running for an hour. The fact that it had been running for two days proved (to me) that at least there was no problem with gears binding nor with the Weight not being big enough, so the issue had to be down to the Escape Train geometry. After many attempts to find that ‘Sweet Spot’ - without success - I left it for another day!

There is some good news about the Lead weight - - - - Mighton Products had been excellent in their dealings so I thought it pertinent that they should know of the delivery issues so I sent detailed feedback purely to make them aware that FedEx has failed in their ‘Next Day’ service. Surprise, surprise – within 2 hours I had a promise of a delivery charge refund which materialized the following day!! This was prior to their knowing whether they could make a successful claim against FedEx. This is not just an inconsequential amount of a couple of pounds, it amounts to ~23% of the cost, nearly £12 so very welcome.

26th - Feb
During my attempts to get the second

(shortened) Pendulum working, I had made changes to the ‘Beat Adjuster’, so, as I anticipated, the clock had stopped after less than 10 minutes. This made me take a more ‘scientific approach’ to the Escape Train geometry. Essentially, with the Pendulum ‘At Rest’ the tips of both the Latch and the Finger should just touch the Escape Tooth and Pin respectively - see the red circles in Fig-223 - so I spent some time this morning re-setting the Beat Adjusting Screws along with the Pivot Adjusting Screws to achieve this configuration.

With just a very slight nudge to the Pendulum Rod, I knew within a few seconds that I’d been successful - the ‘tick’ was in beat !

Now, at last, I could turn my attention to the calibration of the pendulum length. I had wound the Adjusting Nut right to the top of its potential travel and after only half an hour I could see that it was 2 minutes fast so I wound the Nut down about 10 turns and re-set the hands. Again there was a perceptible gain after another half hour so I marked the Adjusting Nut and made an accurate 2 turn adjustment. After 4 hours it was showing about 2½ minutes slow. I’ll leave it for a while before making any further changes but at least it seems that I’m moving toward a working and accurate time-piece in an orderly fashion. More importantly though I think I now know how to set the Escape Train up ‘by eye’.

-----------------------------------------------------------------------------------------------------------
The attached .PDF file includes the posting from Monday.

 

[donwatson]

Thanks JG

 

[J-G]

2nd Mar
It seems that I spoke too soon

Due - I think - to the change in the weather, certainly here it is a few degrees warmer and probably more humid than of late, it seems that the second train main gear has distorted and become ‘oblate’. I only know this because I eventually made a point of marking the 64T Gear whenever the clock stopped, that is every few minutes!

After about twenty or so restarts I was able to notice that there was a preponderance of marks at 180° to each other.

I had by now re-located the clock from my workshop to my office so that I could more frequently check how the timing was progressing without constantly moving to and from the workshop. This meant that I had to dismantle the Pendulum and re-attach it in the new location which meant that the delicate geometry between the latch/finger/pivot was upset so I needed to make small adjustments (I think I need to find a way to ‘fix’ the pivot screws once I have them at the correct position). Because I had been making small - I’m taking less than a ¼ turn of an M3 screw - adjustments to the ‘Beat’, as well as the pivot screw positions when trying to find that ‘Sweet Spot’ again, the tick was no longer ‘In Beat’ so I was fighting well against the odds

In Fig-224 you can see the pencil marks I’d been making each time the clock stopped. They are not all in the same (ish) spot but there is certainly a pattern, which is repeated directly opposite.

This leads me to surmise that the outline is now nearer to that shown as a red line in Fig-225 – somewhat exaggerated of course – as opposed to the concentric black outline. I’ll have to do some more work on the other 64T gears before I build the other clocks but for now I’m resigned to waiting for the weather to settle down again before spending more time in trying to get the first clock fully operational.

3rd Mar
In the meantime, two things have been keeping me occupied.

1 - When needing to make adjustments to the length of the Pendulum I found that I couldn’t remember from one adjustment to the next which way I was

making the change so had to work it out from first principles each time! This made me decide to add marks to the Bob showing [F] and to signify whether the adjustment should be Faster or Slower. I’ve made small Maple buttons engraved with the two letters filled black. (Fig-226) These will be set into recesses in the Bob, either side of the Adjusting Nut.

2 - I’ve decided that I could add a ‘Seconds Sub-Dial’ to the basic design. Had I done this originally, it would probably have been at the 12 o’clock position which might have impacted

upon the design of the Dial but, doing it with hindsight, it will go 50mm lower and be set inside the main Dial. This means that it is slightly more complex in-as-much-as it will need three gears rather than a direct ‘pointer on the end of the Escape Spindle’.

I’ve shown the Gears super-imposed on top of the Dials in Fig-227 and - in the light of the issues I’ve had with respect to the ‘binding’ of the other gears - I’ve changed the tooth profile to provide more clearance. In reality the Gears will all be behind the frame. As silly as it sounds, this ‘complication’ has added 20 components to the total and that’s only counting the ‘dots’ as one!

 

[donwatson]

The weather conditions do seem to have an effect. I would opt for building it where it is going to hang and allow all the materials to acclimatise to the conditions. But as we all know they put a case around the grandfather (and other) clocks to stop the wind/draught affecting the pendulum so it looks as if it will be a long road to get the wood conditioned ??

 

[Yojevol]

This means that it is slightly more complex in-as-much-as it will need three gears rather than a direct ‘pointer on the end of the Escape Spindle’.

More gears = more friction = more weight

Is there any particular reason why the last 2.5 paragraphs appear to be struck through? Is this how they appear to you, @donwatson? They appear OK when viewed on the PDF
Brian

 

[J-G]

The weather conditions do seem to have an effect. I would opt for building it where it is going to hang and allow all the materials to acclimatise to the conditions. But as we all know they put a case around the grandfather (and other) clocks to stop the wind/draught affecting the pendulum so it looks as if it will be a long road to get the wood conditioned ??

I would agree that acclimatisation of materials would be ideal and I do try to keep even raw material as well as part worked stock in an 'inside' environment - my main workshop is directly attached to the house and even my secondary workshop (1m gap between this and the house) is on the same central heating circuit !

One of the clocks is was destined for Essex and another for California so that adds a level of complexity.

 

[J-G]

More gears = more friction = more weight

I take your point Brian but hope that with the new (leaner) tooth profile and the fact that there will be a needle roller and a deep groove ball bearing in the train will mitigate the possibility. It will certainly add to the learning curve!

Is there any particular reason why the last 2.5 paragraphs appear to be struck through? Is this how they appear to you, @donwatson? They appear OK when viewed on the PDF
Brian

Can't explain the strike-out but I've just done an edit and it seems to have sorted it. I certainly didn't select the strike-out intentionally at any time and the icon didn't show 'live' but click [on] and then [off] did the job.

 

[donwatson]

I did see the strikeout originally but it has gone now.