George Daniel's Watchmaking understanding!

amoo.ara

Registered User
Jan 12, 2021
7
2
3
29
Country
I was reading the Daniels' Watchmaking last night and there were a part of it that I can't quit understand. I assure you I did my best to understand it before I write this but it just doesn't make sence to me! if you could help me with that I will readlly appericiate that guys. here's the text:(page 14 &15)

" When using the mandrel plate with dogs for uprighting holes or turning recesses the work must be truly centred. The pump centre of the plate is useful for locating the work but cannot be relied upon to centre precisely. When opening a hole for a jewel to be fitted, or drilling a bridge for alignment with a lower hole, absolute concentricity is necessary to ensure the fitted component is upright. For example, a barrel of 24 mm diameter is to be fitted beneath a bridge 4 mm high and with 0.2 mm clearance. If the bridge is bored only 0.02 mm out of upright, the barrel will have 0.14 mm clearance on one side
and 0.26 mm clearance, almost double, on the opposite side. This would look very unsightly and, if the error were repeated in the centre wheel, the two would probably touch. There is no room for any tolerance in uprighting if the watch is not to become unnecessarily thick."

So my questions are,1. how did he came up with thoes red numbers? (0.14 mm and 0.26 mm)
2. what does it have to do with thickness of the movement?

p.s: It will be super nice if you can draw it for me. I know it's a lot to ask but that would be a great help and I'll be greatful forever.
 

roughbarked

Registered User
Dec 2, 2016
7,992
1,615
113
Western NSW or just this side of the black stump.
Country
Region
I'm not vey good at writing mathematical formulae. So I won't attempt to make myself look any sillier than I am.

However, by thickness of watch becoming unnecessary, he's referring to the fact that an error once multiplied comes out totally overblown.

I'd look further in the text for a reference to the arbor, which seems to be missing from the copy/pasted text. As it is usual for the arbor to have different sized pivots. A barrel also has different diameter's top and bottom.
 
Last edited:

Daniel Keen

NAWCC Member
May 30, 2019
7
2
3
Country
Region
It's because the barrel is cocked to the side. Here's a 4mm tall barrel that's 0.02mm off at the very top. Overall picture first, then zoomed in on the top misalignment and bottom. You'll see that on the left side of the image we've lost 0.06mm of clearance, so you'll end up with 0.14mm of clearance, and on the right, you end up with an extra 0.06mm of clearance, so we end up with 0.26mm of clearance.

The reference to thickness becoming unsightly is because to add slop for these mistakes in misalignment and thus prevent wheels from hitting each other you need more clearance space, and the watch therefore must become thicker.

1656365990855.png

1656365861845.png 1656365972462.png
 

amoo.ara

Registered User
Jan 12, 2021
7
2
3
29
Country
I'm not vey good at writing mathematical formulae. So I won't attempt to make myself look any sillier than I am.

However, by thickness of watch becoming unnecessary, he's referring to the fact that an error once multiplied comes out totally overblown.

I'd look further in the text for a reference to the arbor, which seems to be missing from the copy/pasted text. As it is usual for the arbor to have different sized pivots. A barrel also has different diameter's top and bottom.
Thanks for the answer. I believe it's just a simplified example to show the necessity of work concentricity of work that's why he didn't mentioned the arbor's diameter.
let me give it a try! I drew what I could understand from the text. could you take a look to see if you get the same idea as I did?
what I could understand is that the arbor of the barrel should be 0.2mm thinner than the coresponding hole.(that is what he calls clearance, right:???:?)
for instance if the arbor has a diameter of 1 mm. the hole should be 1+0.2mm, which means 1.2mm in diameter.(0.1mm wider than the arbor from each side!) so if the holes of the bridge and hole of the main plate were not places exacly where they were supposed to, and have a placing error of 0.02mm then you should end up having something like what I drew in the second picture! and my numbers are not similar what Mr.Daniel mentioned. plus, it has nothing to do with the thickness! ANNNNND I don't get the part about center wheel at all!!!! THE TWO WILL TOUCH? Which two? What does he even talk about?

photo_2022-06-28_00-30-24.jpg photo_2022-06-28_00-30-15.jpg
 

Chris Radek

NAWCC Member
Apr 13, 2014
1,191
643
113
Lincoln, NE, USA
timeguy.com
Country
Do not make this too hard.

Imagine a barrel 4mm tall and 12mm in radius, hold the bottom pivot in place and move the top pivot over 0.02mm. The barrel's edge will go up 0.06mm on one side and down 0.06mm on the opposite side. This is because 12/4 is 3; so if you move the top pivot any amount, the edges will move 3x as far.

Now imagine your barrel is only 2mm tall, and everything else is the same. Since 12/2 is 6, now any uprighting error is multipled 6 times at the edges of the barrel instead of 3. Thin watches have to be much more perfect.
 

amoo.ara

Registered User
Jan 12, 2021
7
2
3
29
Country
It's because the barrel is cocked to the side. Here's a 4mm tall barrel that's 0.02mm off at the very top. Overall picture first, then zoomed in on the top misalignment and bottom. You'll see that on the left side of the image we've lost 0.06mm of clearance, so you'll end up with 0.14mm of clearance, and on the right, you end up with an extra 0.06mm of clearance, so we end up with 0.26mm of clearance.

The reference to thickness becoming unsightly is because to add slop for these mistakes in misalignment and thus prevent wheels from hitting each other you need more clearance space, and the watch therefore must become thicker.

View attachment 714622

View attachment 714620 View attachment 714621
It's because the barrel is cocked to the side. Here's a 4mm tall barrel that's 0.02mm off at the very top. Overall picture first, then zoomed in on the top misalignment and bottom. You'll see that on the left side of the image we've lost 0.06mm of clearance, so you'll end up with 0.14mm of clearance, and on the right, you end up with an extra 0.06mm of clearance, so we end up with 0.26mm of clearance.

The reference to thickness becoming unsightly is because to add slop for these mistakes in misalignment and thus prevent wheels from hitting each other you need more clearance space, and the watch therefore must become thicker.

View attachment 714622

View attachment 714620 View attachment 714621
Buddy! I really liked your drawings, it helped a lot. Thank you
 
  • Like
Reactions: Daniel Keen

amoo.ara

Registered User
Jan 12, 2021
7
2
3
29
Country
Do not make this too hard.

Imagine a barrel 4mm tall and 12mm in radius, hold the bottom pivot in place and move the top pivot over 0.02mm. The barrel's edge will go up 0.06mm on one side and down 0.06mm on the opposite side. This is because 12/4 is 3; so if you move the top pivot any amount, the edges will move 3x as far.

Now imagine your barrel is only 2mm tall, and everything else is the same. Since 12/2 is 6, now any uprighting error is multipled 6 times at the edges of the barrel instead of 3. Thin watches have to be much more perfect.
I wish I had a geometry teacher as good as you! even half as good as you could've changed the a lot.
 
  • Like
Reactions: roughbarked

dave-b

Registered User
Jul 28, 2010
101
10
18
Country
If the pump centre cannot be relied on, how is perfect concentricity achieved?
 

Chris Radek

NAWCC Member
Apr 13, 2014
1,191
643
113
Lincoln, NE, USA
timeguy.com
Country
If the pump centre cannot be relied on, how is perfect concentricity achieved?
A wobble stick. (This shows a fancy metal one, but I just use pegwood shaped with a razor blade)

 

Dr. Jon

Moderator
NAWCC Member
Dec 14, 2001
7,346
1,690
113
New Hampshire
Country
Region
The pump center on a mandrel plate is bored into a plate which is then set into a collet on most lathes. That is rtoom for a error build.

The usual way to verify centration is too put a sharpened peg wood stick in the hole and support so th tother end is near the drill tailstock. ( A wobblestick as per the above post) If out of true the stick will swing in a cone as you rotate the head stock. The other method is to use a dial indicator either on the hole or a knnown round rod firmly inserted in the hole.
 

dave-b

Registered User
Jul 28, 2010
101
10
18
Country
Thanks. I had completely forgotten wobble sticks. ( have a dedicated mandrel lathe which seems very accurate .I must do some comparative tests)
 

gmorse

NAWCC Member
Jan 7, 2011
14,285
3,400
113
Breamore, Hampshire, UK
Country
Region
Hi Bruce,

George seldom gave absolutely full explanations, he tended to lead you down the path but left some details for you to work out for yourself; he expected his readers to think!

Regards,

Graham
 

gmorse

NAWCC Member
Jan 7, 2011
14,285
3,400
113
Breamore, Hampshire, UK
Country
Region
Hi Dave,
I have a dedicated mandrel lathe which seems very accurate .I must do some comparative tests
I think these dedicated headstocks do tend to be more accurate, I've been unable to detect any runout on the pump centre in my Lorch Schmidt, although I always double check with a wobble stick.

Regards,

Graham
 

Betzel

NAWCC Member
Dec 1, 2010
819
182
43
Country
Region
To me, our target of perfection is only as good as our ability to understand and then repeatedly measure anything. The better/more you see, the worse it gets.

So, with a well-cut pump point a center things look fine to the eye (esp. on those made with pride and well maintained), but deviations may show with a wobble stick. I've never used a microscope tailstock, but if it were of high quality and resolution, you would then be able to measure even better and likely see errors not easy to measure with a wobble-stick, and so on down the line. Maybe a laser is next? Same thing with straight distances and angles (a perfect 4% taper is not easy to cut, grind or measure, especially if it's short!)

Fortunately for us, George often suggested the methods in his books were "good enough" for the work he was interested in, which I think was mainly pocket watches. And this is reassuring so we can focus on the real work after getting good with better tools and tricks. But I'll bet he had some fancy-pants tools to carry out co-axial escapement research and development. Fortunately, we're not operating at that level?

Great thread, folks.
 
  • Like
Reactions: gmorse

praezis

Registered User
Feb 11, 2008
576
75
28
Germany
Country
Region
what I could understand is that the arbor of the barrel should be 0.2mm thinner than the coresponding hole.(that is what he calls clearance, right:???:?)
Not at all! You assume sideshake, which will be maximum 0.02 mm here.
Daniels' 0.2 mm is the space between barrel and main plate / barrel bridge.

I found a wobble stick much too inaccurate for centering. I use a dial indicator mounted in the tailstock, with different interchangeable inserts/pins

Frank
 
  • Like
Reactions: Jerry Kieffer

karlmansson

Registered User
Apr 20, 2013
3,029
313
83
Linköping, Sweden
Country
I'm suspecting that this has been adequately explained already but in the spirit of making sure:
Daniels is talking about barrel tilt. By boring one part out of upright from the other you will have created an axis that is out of upright and the barrel will need to have additional headspace as not to be obstructed. Hence the talk about the thickness of the movement.

Apart from the wobble stick method (which I believe is described on the same page that you quoted, 14) you can also use a centering microscope or a loupe with a good, stationary, visual reference. The idea is to have a stationary reference to compare the circumference of the rotating hole to. Nothing wrong with the wobble stick method but if you are using a steel wobble stick there is a risk with being too heavy handed with it as you may burnish the setting you are trying to center and losing your reference edge in the process.

Regards
Karl
 

Forum statistics

Threads
176,402
Messages
1,544,018
Members
53,286
Latest member
ferkor
Encyclopedia Pages
1,064
Total wiki contributions
3,031
Last update
-