Last modified: 2023-09-15 19:06:28
< 2023-09-14 2023-09-16 >I've been thinking a bit more about putting a "lever" on the escapement. I think I'd like to see some sort of prototype of it. Whatever is easiest out of 3d-printed or simulation.
I'm also keen to try and have a harder look at the Roskopf escapement.
Jacques Favre has this video: https://www.youtube.com/watch?v=aS52AALli2U
It looks like it is not the same as the simple lever idea I want to try, it has all the complex interactions of locking, banking pins, guard pin, etc.
"In the original watch the pins are 0.2mm in diameter" - good fact.
The difference between the English lever and the Swiss lever is whether the lever pivot is inline with the escape and balance wheel (Swiss) or whether it is at 90 degrees (English): http://horology-student.org/themes/types-of-escapements/#english
For making a lot of the parts, I'm going to want to start on the lathe and then move to the router table, so I will need a way to hold the superglue arbor vertical. I have 2 ideas:
Or maybe just get long enough bolts to hold a chuck down with strap clamps?? Let's go with that for now. Especially the collet chuck won't need super long bolts, will just need a spacer to make the mounting studs clear the table.
Let's try out the process of starting on the lathe, just to practice, so I'll want to:
Well, that didn't go to plan. Given that I was going to have to make a disc out of 3mm plate anyway, I couldn't be bothered transferring it to the lathe and back, so I just machined the whole thing on the router.
It went fine except the lead-in/out gouged the work, idiotic mistake, should have noticed that in FreeCAD:
Next up: make some 1mm pallet pins!
Before adding pallet pins, I tried sticking some tape to the light side of the balance wheel to see how much weight it needs to be balanced. Looks like about 0.14g almost does it. One M1.6 screw is about 0.06g, so 3 or 4 them should be enough. I plan to try to use the 4th axis to drill some holes around the circumference to put screws in.
I turned up a couple of 1mm pins and hammered them in:
I did try pressing one in with the vice but it just bent. Hammering is the way to go. I need a better method for making the pins. These are tapered to ensure a good fit, but that means the bit that sticks out is also tapered. Maybe the plan is to make a tapered section that hammers into the hole, but make the pin narrower so that it can be straight? Or maybe taper it from the other end so that the only tapered part is the part in the hole?
Also I really need to harden the metal a bit because it bends too easily while turning. Several times I took too deep a cut and had to hammer it back into shape. Maybe I should just buy blued pivot steel instead of wasting all my 4mm silver steel?
Wacky idea: what if I make a quick 3d-printed frame and use the 1.75mm wire as pivots? Will it work? That should be enough to let me try to turn the escape wheel by hand, no?
Then perhaps I can look at putting M1.6 holes around the perimeter of the balance wheel?
The 1.75mm wire is too small. Maybe try with 2mm drills?
Pretty much works, a bit: https://www.youtube.com/watch?v=NaEK9nB4ebA
I think the wobbly pivots & high friction are making it bad. Also maybe the very-sharp teeth on the escape wheel are making it catch. Maybe because the pins are slightly bent?
It looks like it'll want to run very fast. Will need a lot more mass on the balance wheel, especially after the hairspring is added. How much denser is brass than aluminium? I should add densities to jesref. (About 3x denser).
Or maybe the "lever" idea is required to make it tick more slowly.
Next I'm going to try to make some better pivots.
When turning the pivots, if you rub the grey scotchbrite stuff against the end it polishes it quite nicely. At least, it's considerably better than doing nothing.
I managed to bend one of the pivots trying to hammer it into the escape wheel. Definitely need to be using hardened steel for this. Also I think 1 degree of taper is too much. However, I have created these:
I used Loctite 603 to glue them to the shafts, which is apparently a big no-no in watchmaking, but I'm not a real watchmaker so I don't care.
Anyway, this feels like a good milestone! For the first time, I have made some watch parts and mounted them on shafts. I need a frame that I can mount these on now, maybe metal?
Notes for next time I make shafts:
Maybe I should make a dedicated notes file with a "playbook" for each type of operation I need to do, that I can try to follow and keep updated as I find improvements.
It's in watchmaking.md.
I think I want a way to publish these notes that makes them easier to read (and share?) than the github view. Some sort of cron that pulls the github repo and rebuilds a static-site if anything has changed.
There are 2 types of file:
They can be identified because daily notes are yyyymmdd.md and topic notes are not.
Sometimes I paste imagebin links in the text, I probably want these expanded to show the actual image, but otherwise just markdown-to-html, a chronological index page, and some sort of index of topic pages (maybe just a navbar for now, until there are too many? in future could make it hierarchical).
I wrote a static site generator similar to the one for the blog. It could maybe do with a search function (and for that matter, maybe the blog could too). I don't know whether the best way is to generate an index in a javascript file which is loaded on-demand, or maybe a bunch of indexes named by the first few characters of the search term, or maybe just link to a DuckDuckGo search.
https://incoherency.co.uk/notes/
Need to order supplies to build the ceiling. I'm debating whether to construct a wooden frame, which is faff and expense but will simplify the fitting of the plasterboard, or whether to attach the plasterboard directly to the aluminium, which will be much more annoying than attaching it to wood.
I think just attach the plasterboard to the aluminium, get Emma to help, and use technique to make it easy. It would probably be easier to fit a larger number of smaller sheets instead of a small number of full-size sheets. Maybe just cut them into quarters once they arrive, if we find fitting full sheets too difficult.
Need to buy:
Came to £105 for the plasterboard and OSB from Wickes, and £30 for the screws from Amazon. Not too bad.
What's the plan for tomorrow then?