One step closer...

One day I intend to make a watch.
I’m still a long way off, but tonight I made a good step in the right direction :)

A large part of a timepiece is of course the wheels and pinions. I’ve been contemplating how to cut them without bankrupting myself buying the cutters.

The Sunderland gear planer uses a fairly simple cutter for racks, and generates the curves of the spur gears by coordinating the cutting and a rotation, so the gear blank rolls against the rack as though it already had th teeth.

I have several old engineering texts with some of the relevant details in:

https://i.postimg.cc/TYFhHPcZ/D3-D84...19443-FAE8.jpg

Of course I don’t have a gear planer, and generally they are for much larger gears, but I can use the technique with a little bit of lateral thinking.

First make a rack form cutter, simple lathe work.

https://i.postimg.cc/gkT6FsGt/0-DC70...-E77-DFEFC.jpg

Then wack it into an indexer on the mill

https://i.postimg.cc/bvQGnjk4/CD60-E...735-ACA4-C.jpg

And create some flutes for the cutting edges.

https://i.postimg.cc/mDLrF1bF/3009-D...EC4-FA16-F.jpg

Then handily I have recently bought a kiln (for enamel) which I can use to Heat treat. So much easier than a blowtorch.

https://i.postimg.cc/MpffWBT6/608-F8...C2881-BCC6.jpg

I decided to keep the tooth count smallish for my first gear - 30 teeth /12degrees each. That also means I can use the simple divider. To assist my little brain I wrote down all the teeth angles so I could tick them off as I go.

https://i.postimg.cc/x1mqrSxq/232-BA...-DE84970-B.jpg

Then it’s a question of put on the right cut and index /cut / repeat...

https://i.postimg.cc/KjY4Ckwy/D3-F63...B7-CB063-D.jpg

Once the first pass is done I rotated the indexer by 6 degrees, moved the cutter by an equivalent linear amount and cut it again. This generates the tooth curves.

The end result is gear shaped:

https://i.postimg.cc/sDvvcW5Y/169-DB...71-C81-D11.jpg

And meshes with another 48dp gear nicely.

https://i.postimg.cc/9ftMMYjh/036036...5162387463.jpg

I got a beautiful pattern when parting the gear blank off the stub it was machined as part of.

https://i.postimg.cc/g0PjmBQs/A4-ED5...78-E98-C94.jpg

There is more work to do, but I think the method will work without too much effort.

Dave

Great work Dave! Look forward to future progress. Meanwhile, that Porsche . . . ;-)

Great work. Thanks for sharing....and keep us posted on the rest of your "journey". [emoji846]

Sent from my G8441 using TZ-UK mobile app

Had to make another cutter as the first one is too short to make a 144 tooth gear.
I printed out the 144 tooth gear to compare the lengths and clearance with:

https://i.postimg.cc/Dz42jQ3H/56-FB3...0-A1-C2466.jpg

Then I milled some flutes and a bit more clearance on the shank.

https://i.postimg.cc/wvPDKQ1N/EEB4-D...6472965-CA.jpg

Bit of a heat treat and I knocked some steel into the shape of a pinion.

https://i.postimg.cc/zvzvgVLL/IMG-6012.jpg

Then I whacked the bar into the Lathe and proceeded to put a hole in the middle and part off a few pinions for closer examination.

https://i.postimg.cc/wjvNrbjt/IMG-6017.jpg

They need a good debur (I need to sort a better parting tool I think) but the tooth form is quite reasonable.
https://i.postimg.cc/qq4hx1Fv/IMG-E6018.jpg

Dave

Good to see some some progress Dave, keep at it!

Dave - can't believe there havent been more comments on this... epic work! I'd love to see a few videos of the milling and lathe processes. If you ever get a chance to record them and upload them to YouTube it would be great! Looking forward to future updates!

Ive made some wheels.
First trepan out some blanks, which avoids making loads of swarf just to get a roughly right sized blank.

https://i.postimg.cc/Dz2M191V/IMG-6046.jpg

2 cut out without a hitch Happy face :)

https://i.postimg.cc/ydxpb0fv/IMG-6048.jpg

Because they are only roughly the right diameter I made a stack of Gage Blocks to allow me to size them to within a very small tolerance

https://i.postimg.cc/zfG1qbtR/IMG-6073.jpg


Turn to correct outer diameter in the lathe, whilst on the arbour they will get gear cut on.

https://i.postimg.cc/F15Pz8hB/IMG-6069.jpg

Cut some teeth
https://i.postimg.cc/pVhR9j2f/IMG-6082.jpg

And repeat x144. Took less time than you might think

https://i.postimg.cc/dVBYMmBY/IMG-6084.jpg

Now the blanks have teeth its time to give them some spokes.
so I marked out 3 spokes:
https://i.postimg.cc/Pq9GjtCW/IMG-6105.jpg

cutting a long thin slot is fraught, and usually ends up with a broken tiny endmill, so I went the easy route and chain drilled out the scrap.

https://i.postimg.cc/28msm9SK/IMG-6106.jpg

https://i.postimg.cc/WzVHNtTk/IMG-6108.jpg

Then some filing to clean up the drill half holes

https://i.postimg.cc/httNQJx8/IMG-6117.jpg

And a pair of wheels is done enough to make the next bit.
https://i.postimg.cc/yNJ5nwLk/IMG-6124.jpg

Ill do 'proper' finishing when the rest of the movement is more complete.

I'm learning quite a lot about things that will help when I get to watch sizes, although I'm still not doing any work on the Porsche...

Dave

Well I think you're bonkers Dave S !
In an eccentric way of course.
I think you have to be to make a watch by hand.
I think of people like George Daniels, Kari Voutileinen, the late Derek Pratt and of course AL Breguet. I know how much work making even one part can be and wish you all success.

Sent from my SM-G900F using Tapatalk

Great stuff Dave! Now, about that Porsche . . . ;-)

Brendan, I’ll take that as a complement :)
I have to make a couple of timepieces - one for each child...
New book on gears on order - these are involute tooth form rather than the traditional cycloid form, but I haven’t yet found a complete set of writings which have convinced the engineer in me that involute really is a no no.

Dave

Quote:

---

Originally Posted by Dave S

Brendan, I’ll take that as a complement :)
I have to make a couple of timepieces - one for each child...
New book on gears on order - these are involute tooth form rather than the traditional cycloid form, but I haven’t yet found a complete set of writings which have convinced the engineer in me that involute really is a no no.

Dave

---

Of course it's a compliment !
Such skill and patience is required.
It's a while since I made a wheel !
From what I remember the involute gear form is used by most heavy industries, gearboxes etc, and will work well in both directions but sometimes require lubrication.
With cyclical gears there is less friction but depth is more critical and devices turn best in only one direction due to tooth contact interfaces.
*You've got to make three ! I want one too.

Sent from my SM-G900F using Tapatalk

Great story with some lovely machining

Well done that man

Bit more progress, so a quick update.
I hacked out a couple of more wheel blanks, this time for 96 and 90 tooth wheels.

https://i.postimg.cc/HW9xcN7M/IMG-6122.jpg

Then bunged them in in the mill and cut some teeth in the same manner as for the 144 tooth wheels, of course cut one at a time, as they have different tooth counts...

https://i.postimg.cc/j5zjhggG/IMG-6198.jpg

https://i.postimg.cc/50c0Ld3B/IMG-6270.jpg

Now I have 4 wheels, the smaller 2 of which need crossing out for spokes:

https://i.postimg.cc/MpFXZjfr/IMG-6274.jpg

The layout for the hands will be in a classic Regulator style - Large central minutes, with smaller hours and seconds, similar to this picture stolen form the internet:

https://i.postimg.cc/6Q7mGFjf/217024...1493214397.jpg

So the wheels end up laid out something like this:

https://i.postimg.cc/tJkCK1Hk/IMG-6272.jpg

The hours are on the lower large 144 wheel, the minutes on the middle wheel, and the seconds end up on the escape wheel shaft.

The steel for the pinions and shafts has appeared, so I'll be cutting them shortly, along with escape wheel and then plates.

I've also got some reading material, about 1/2 way through it so far. Lots of maths and formulas, so needs to be read with attention.

https://i.postimg.cc/V6MJVL6Z/IMG-6212.jpg

Dave

Love the technical jargon Dave, "hacked" and "bunged"! Good to see more progress.

Soon you will be able to build one of these??

https://live.staticflickr.com/65535/...5841f75f_c.jpg

Following a lot of reading about aspects of gear manufacture I've started on the Pinions.

When I get to Watch sized parts I think I might want to make a slightly smaller setup for cutting.
(Anvil for scale ;) )

https://i.postimg.cc/15fNdf7V/IMG-6345.jpg

Cutting is much as before

https://i.postimg.cc/c428G0P9/IMG-6346.jpg

Involute Pinions with low tooth counts suffer from undercutting, where the root of the tooth becomes thinner.
This is often quoted as why you cant use involute gears in a time piece - cycloidal profiles dont suffer from this as the root flanks are radial.
To counter the undercutting it is common in industial gearing to use a concept known as profile shift.
This means that the involute tooth is still involute in shape, but a different part of the involute curve.
'Form' gear cutters - as generally used in clock making - dont allow profile shifting.
However with a rackform generation process (such as hobbing or planing) all you do is make the blank bigger, and move the cutter out a bit. (Danger - slight oversimplification)

Doing this does mess with the actual pressure angle, but I have more thinking to do about the consequences of that.
In the mean time here are some pinion teeth cut with a +0.25mm (approx) profile shift.
I havent taken them out of the mill yet, as there are a few things I want to try whilst the setup is still in place.
You can see the effect of the shift to give a "pointy" tooth shape with no undercutting.

https://i.postimg.cc/pV0FkDCt/IMG-6349.jpg

compared with the same pinion without profile shift

https://i.postimg.cc/qqZ6RTq7/IMG-6190.jpg

Either pinion will roll correctly against the previously cut wheels, but the shifted one has a different center distance (as its physically larger).
That matters if the gears have to fit into an existing plate, but as I'll depth the train anyway it will 'just' get accounted for.


Dave

The cycloidal form is used in clockmaking/watchmaking.
Out of curiosity why did you decide on involute?

There are a few reasons, but mostly it boils down to being bloody minded and wanting to try something that is not normal ;)
Whilst traditionally the cycloid tooth form is used I have failed to find any actual evidence that it is really superior to the involute form - both tooth forms produce a rolling action, both are theoretically “good”.
I have found a lot of “hearsay” about various aspects, including how involute gear trains don’t work for gearing up ( how does the gear know?), and that you can’t cut low tooth count pinions. There is some truth to the latter - with an off the shelf commercial form cutter you can’t, but by profile shifting it’s actually quite common in industrial practices.

Another aspect is that I can make pretty much any tooling I need to - my shop is well equipped. I have wanted to try this method of a single (rack form) cutter for all gears for a while, and making that tool for involute gears is trivial. Being an engineer I’ll take the path of least resistance the solutions are generally more elegant :)

As I said I intend to progress to watch sized timepieces, and being able to make my own tools, and know how to run the process are (IMO) useful bits of knowledge.

Dave

Cycloid runs with lower friction in lower torque situations.

Do you have a source for that statement?
Dave

Pretty sure it will be in Theory of Horology by WOSTEP, but it is also in my gears book but can't remember the title. Will look it up later when I get home and post it here, its not an expensive book and worth it.
Have you read wheel and pinion cutting in horology? Well worth the money that book.

I have Wilds Wheel and Pinion cutting, though it’s a few years since I read it. I’ll have another look, but IIRC it just makes a statement to the effect of gear trains for clocks are cycloidal, but involute is sometimes used in the winding works. I don’t recall anything more scientific than that.
I’m currently still working through WO Davis.
Would like to find some actual data showing the comparison of friction / torque for involute and cycloid gear trains - not seen any proper data for that yet.

Dave

I'm not one side or the other, I suggested the wheel and pinion book just because it's useful for this, rather than one profile or another being better.

The book is A W Marshall - gear wheels simply explained. But by its title it's only a paragraph, no real data to back it up.
Would be great to make two trains of both forms and test them for power reserve, see which lasts the longest.

Let me make one train first, then maybe I should make another and do the research directly...
Of course that would mean buying the cutters which the whole point was to avoid....

I'm not sure there is one right answer, inevitably there are a set of trade offs in almost any engineered thing.

Dave

Indeed, involute is also easier to depth. I would like to be able to find more information on the other profiles used now like microgear. Very hard to find any information at all on that profile.

Fascinating post Dave, Thank you for sharing your work.

I have the suspicion that this is the sort of thing I wanted to study while doing mechanical engineering, not attempting to get my head around La Place transforms.

Looking forward to the next installment, don't worry about the Porsche, it can wait ;-)

Been busy laying a hedgerow, so this took a break for a bit, but now that’s done it’s time to get on with it.

I managed to mis-index the pinions whilst trying at few extra things, and as a result knackered them.
So tonight I knocked out another “stick”.

https://i.postimg.cc/P5FY5cDJ/B20-B6...F1-F24-F41.jpg

https://i.postimg.cc/9FCG8yVm/36436-...-A2-F281-C.jpg

I haven’t made a depthing tool yet, and that would require finished wheel/pinion sets anyway, so I bunged the stick in my little unimat and found a bolt to act as the wheel shaft.

The gears and pinion mesh nicely enough given the roughness of the jig.

Not sure how to post a video here, so here is a link to a tweet with a couple of seconds in:
https://twitter.com/solutionsbydave/...382662657?s=21

Dave

Fascinating Dave. Look forward to future updates.

dunk

Get that hedgerow finished! We need to see more of this wonderful machining.


Sent from my iPad using TZ-UK mobile app

A long stick of pinion is not much use, so I parted three pinon shaped pieces off it.

https://i.postimg.cc/GhBVQZpC/IMG-6830.jpg

Then I could try a rough positioning of the the train. It all fits and I can get pivot shafts thorugh the pinons without fouling on the wheels :)

https://i.postimg.cc/FKXM1DNQ/IMG-6827.jpg

Im not totally certain what Im doing for the bearings yet, but I think the shafts will be 4mm diameter.
The pinions will be fitted to the shafts, with the wheels held on by collets.
So I threw some holes roughly in the centre of the pinons:

https://i.postimg.cc/tgzMJbBh/IMG-6831.jpg

Ill then bore these out on the lathe so the teeth are concentric with the shaft.
To do that Im going to use a soft collet, bored in place, which will hold the outer of the pinion.
Luckily in my pile of brass bits I found a piece of hex stock which already had a hole in the end slightly under the size I need.

https://i.postimg.cc/k5CH7GvJ/IMG-6833.jpg

To make this larger I need to use a boring bar. I dont have one that small, but I do have HSS tool blanks and a Dremel :)

https://i.postimg.cc/gkf5BNw6/IMG-6834.jpg

After I finished lopping off all the bits that didnt look like a tiny boring bar I ended up with this:

https://i.postimg.cc/zBXPRQQ4/IMG-6835.jpg
https://i.postimg.cc/t4Wf0WPv/IMG-6837.jpg

I'll use the same bar to open up the pinion bores, hence its a bit smaller than would be needed just to make the collet.

Lathe tools need to be on centre to cut properly, and with small ones there is even less margin for error.
A bit of a hack way to set this up is to turn a pointy end on a piece of scrap in the lathe:

https://i.postimg.cc/qBDW2k4w/IMG-6838.jpg

then without removing it you can use the point to align the tool.

https://i.postimg.cc/J0Fg8QPv/IMG-6839.jpg

I thought about doing this in my smaller lathes, but I really like my big lathe.
It has power feeds, which are helpful in not bumping the tool and breaking it.
I think its a bit overkill for such a small hole though...

https://i.postimg.cc/KjC9w8Yy/IMG-6840.jpg

A few action shots of the collet being bored.

https://i.postimg.cc/hjyZ8rWv/IMG-6841.jpg
https://i.postimg.cc/pXnc8z0V/IMG-6843.jpg

The tool left a nice surface finish, and the pinon is a snug fit, so the boring of its centre hole should be straightforward (probably cursed that now...

https://i.postimg.cc/YCKnx0V6/IMG-6844.jpg

Dave

With the pinon pressed into the soft collet next thing is to bore it out.
Very small, light cuts needed, as the grip on the pinon is not very secure.

After boring to correct the concentricity and bring very nearly to the right size I used a 4mm reamer to finish the hole to size.

https://i.postimg.cc/FKGmTS1n/IMG-6848.jpg

I could have bored it completely, but I dont have a good way of measuring holes this small, so it would be a lot of trial and error as I snuck up on the dimension.
Not really best practice.
Just drilling and reaming would not necessarily put the hole in the middle - drills are bendy and will follow the already drilled pilot hole. Hence bore first, and then size.

https://i.postimg.cc/QNvBQ22K/IMG-6849.jpg

To get it out of the press fit collet I had drilled and tapped the back of the collet to take an M6 bolt before getting a pinon stuck in it.
The bolt is inserted with a very long allen key through the lathe spindle, and then tightening it pushes the pinion out.
Its almost like I have had a thing stuck in a blind bore before ;)

3 pinons done.

https://i.postimg.cc/hGGPWMFq/IMG-6862.jpg

The wheels also need a little truing up. They were cut on a mandrel that was a tight fit into the central bore, but the machining process can force the wheel / setup to deflect slightly.
To true these up I used soft jaws in one of my 3 jaw chucks.
The jaws a put in and tightened up onto a spacer to remove any play, then a shallow bore of the correct size is made in them.

https://i.postimg.cc/TY8wLrbr/IMG-6875.jpg

Then the spacer piece is removed and the wheel placed into the bored and tightened. Done correctly this means there is no run out on the outer edge of the wheel, adn now the bore can be machined concentric with the teeth.

https://i.postimg.cc/FHNHGNDz/IMG-6876.jpg

I did the same (with a smaller chuck) for the 2 smaller wheels.

https://i.postimg.cc/L8D6bwbT/IMG-6877.jpg

This vimeo link is to a short clip showing how beautifully concentric the pinions are.
It also shows I need to clean up the teeth before hardening them...

https://vimeo.com/394920913

More to come, but I've not uploaded the images to anywhere yet.

Dave

I've not really got anything constructive to add but I just wanted to say what a great thread this is!

Glad people are enjoying it.
I find it nice to share this stuff - my family has got so used to the impossible just taking a little longer than usual that they look bored when I share.
I really will get on with the Porsche one day...

Anyway, the train is progressing nicely, but without an escapement it will just run away.

It always starts with a circle. This one is for the escape wheel.

https://i.postimg.cc/7Zs4KRbG/IMG-6896.jpg

It might be a little bit to big - I'm winging it a bit as I cut the wheels to a different size than the original train design.
I wanted to make them smaller as its more 'watch like' and ultimately the aim is to cut watch sized things.
These are much easier to find when they get dropped on the floor though :)
The end result is that this escape wheel might have a shaft axis that interferes with the next wheel in the train.
I'll find out soon, then I might be re making it smaller...

Bit of sawing at the bench

https://i.postimg.cc/qvM0XrBH/IMG-6897.jpg

The fine saw blade wastes minimal material.

https://i.postimg.cc/hvBRJWZw/IMG-6898.jpg

And with care the blank is quite circular, and has square edges.

https://i.postimg.cc/50vMNNgL/IMG-6899.jpg

To make the escape wheel teeth I need a cutter.
This rough sketch shows the cutter shape.
It will cut the sloping backs of the teeth directly, with the front faces cut at an angle by offsetting the cutter from the centre of the blanks rotation.
Its harder to write than to show with some pics, so I'll do that when I get there.

https://i.postimg.cc/05D89ZXF/IMG-6900.jpg

Some years ago I needed to measure a few angles to more accuracy than a maths set protractor would allow, so I aquired a chinese copy of a russian tool that is specifically made for that.
Handily I can use it now to layout the angles on my tool blank.

https://i.postimg.cc/7ZxPs7Vc/IMG-6903.jpg

Scribing HSS is tricky, but a pointy piece of carbide makes enough of a make for the job, after a light coating of layout dye.

https://i.postimg.cc/rs1cgFtM/IMG-6904.jpg

Then some precision dremelling removes the bits that the not required.

https://i.postimg.cc/1z8Q2KNW/IMG-6905.jpg
https://i.postimg.cc/HLFgg7wT/IMG-6907.jpg

I have a little clean up to do - turns out I cut the angle at 26.5 degrees, which isnt to shabby for a freehand job.

Dave

Great work Dave!

I like the idea, execution and explanation...

Poslano sa mog FRD-L09 koriste?i Tapatalk

In a minor cockup I made the tool from 5/16th HSS. Turns out that the boring bar holder I indented to use it with would work with 1/4", but 5/16 puts the cutting edge in the wrong place.
*Sigh* Time to make some *more* tools.

Rooted about in the scrap drawers until I found a suitable looking chunk. I had already made one end round for a long abandoned project, so it was ideal.
After cleaning it up I painted on some blue

https://i.postimg.cc/MH6RhY3x/IMG-6983.jpg

Did some layout and set it up in the mill.

https://i.postimg.cc/sxHBbCZ9/IMG-6984.jpg

Then proceeded to make some chips. :)

https://i.postimg.cc/154VwXLW/IMG-6987.jpg

I took the smallest of skim cuts to true up the outer surface to the slot, for a later setup and checked the tool fits :)

https://i.postimg.cc/Kj8k9848/IMG-6988.jpg

Then setup for the retaining bolts, using the previously trued up surfaces.

https://i.postimg.cc/jSwN02cW/IMG-6989.jpg

Knocked a couple of M6 threaded holes into the side.

https://i.postimg.cc/XNsyrdDG/IMG-6991.jpg

https://i.postimg.cc/7YcGVD4q/IMG-6994.jpg

https://i.postimg.cc/cHZ8nmSY/IMG-6995.jpg

And checked the tool clamps in securely.

https://i.postimg.cc/0QfKdwk4/IMG-6996.jpg

A little work with a file to ensure the corner is undercut, so the tool will sit square against the side wall,

https://i.postimg.cc/XJQXy6xS/IMG-6999.jpg

Then a quick deburr with a file and a not totally pretty, but functional tool holder is ready.

https://i.postimg.cc/76xf6Z3n/IMG-7001.jpg

Stay tuned for the next exciting episode, where Dave discovers he has setup the tool for LH cutting and needs to crawl round the mill to change the spindle rotation....

Dave

Great stuff Dave! Keep at it. :pig:

Cutter holder made, now its time to sort the blank ready for machining.
Much the same as for the wheels, knock a hole in the middle, mount on the mandrel and true up in the lathe

https://i.postimg.cc/x1vVLFTr/EA85-F...-F7529-CD6.jpg

Back at the mill the tool holder and tool get locked into the spindle.

https://i.postimg.cc/52SNftzF/25-FA1...7-A62886-E.jpg

To get an angled cut on the faces of the teeth one side uses the 26 degree face of the cutter, and the other is caused by not cutting the tooth at the centre of the blank.
In this case the offset is 0.098" - 2.489mm on the tip that will do the 'horizontal cutting.
To set that up I first made a set of gage blocks that were the height of the centre of the blank. This gave me a good measure to base my offset from.
Then I made up a second set with the height offset.

https://i.postimg.cc/g2W29hxW/63178-...74424-DCEF.jpg

Its then fairly simple to line up the cutting tip with the top of the stack

https://i.postimg.cc/RVfZbdVd/FF0-E1...FA9-B8-FAA.jpg

https://i.postimg.cc/j2pB7ypd/68-BD1...260-C80-BE.jpg

This, when you are all set up and ready to go is when you spot that the cutter is in the way that means the spindle has to rotate in reverse...
I deliberately set up the cut this way to push the mandrel into the rotary table, but got my rotation directions wrong.
The mill has a spindle reverse, and indeed 2 speeds for the motor.

But the switchgear is all mounted at the back!

To get there in my compact workshop involves moving the hacksaw, then crawling under the table and round the corner.
Once that was done its time to cut :)

https://i.postimg.cc/59X081hZ/22-C2-...5-FFFD9-C8.jpg


Slowly an escape wheel starts to appear.

https://i.postimg.cc/Y2TS3Zkj/15-C9-...FD73-F87-A.jpg

https://i.postimg.cc/4344yTv2/C882-F...7-B2-B8-B3.jpg

I haven't done the final cut yet - it needs a little bit of measuring first and I haven't had time with the other shenanigans going on.

Dave