How Do They DO This?

[John Buffum]

I've been following amateur watchmakers on Youtube. A thought occurred to me. A watchmaker was working on, I think, a lady's Seiko watch. The case was less than 7/8" wide. The face was held on by two screws that bit on two posts welded to the back of the face. To the naked eye of a 74 year old, these screws are INVISIBLE.!.! Under the microscope, the head is a perfect cylinder. The slot is perfectly centered, squared, and even. The head has a mirror finish. The threads are PERFECT, and continue to the head. They were obviously ground, not cut. The head of the screw (I'm guessing) was less than 1/64" in diameter.

The smallest screw I've made was to replace a rusted out screw holding the windage mechanism of a Lyman 57 rear sight. Though the original had some odd thread, I replaced with a 2-56. Using the tap and die for this was an experience. All came out well, and you can tell the difference, only if you disassemble.

Still, how do watch manufacturers DO this? Without a microscope, you can't even really SEE this screw. Work holding?

[darryl]

I guess you'd need a REAL lathe for that kind of work

[J Tiers]

The watch companies had equipment that turned out those screws by the tens of thousands. I don't think they ground any of them, but it's possible they did. A lot of them are polished-looking.

There used to be some videos on youtube that showed the tiny automatic machines that made the screws. They were about the size of an older typewriter. About 5 to 10 seconds to make a screw.

Not sure about the slot, I'd bet they had a vibratory feeder and another machine to do that, if the screw making machine didn't do it to begin with. You can make a lot of perfectly aligned slots if you set the machine up right.

[portlandRon]

Here is how it's done: https://www.youtube.com/watch?v=NB6EL8Lq7qw

[eKretz]

Looks like thread milling to me.

[jmm03]

but how did they do it in the old days??? How did they grind a thread tool that small accurately?? Jim

[Lew Hartswick]

Looks like thread milling to me.

I haven't made any of those tiny ones but did tap quite a bunch of 0 -80 back in the dim dark ages when I was still working in the electronic business. haven't done any smaller than 2-56 while in the machining field.
...lew...

[J Tiers]

That's now.

Here's the old way. The video is not very good as far as showing the machine, but it's all I could find. A year or so back, I saw some much better videos.

https://www.youtube.com/watch?v=YCmnUP5gx78

[reggie_obe]

Here is how it's done: https://www.youtube.com/watch?v=NB6EL8Lq7qw

Was going to post, talk to a guy who ran a screw machine shop, but portlandRon beat me to it with a YouToob link.

[DR]

Tiny screws and related small parts were made in the old days on Swiss screw machines. They're called "Swiss" probably because they were developed in Switzerland for small part turning. Pre-CNC they operated on cams to feed the stock and the tools. They're also known as sliding headstock machines.

I had an old Tornos Swiss machine with a maximum bar capacity of 1/2" . All the diameter setting of tools used micrometer-like dials capable adjusting to tenths accuracy. My machine was heavy, I'd guess 3 to 4 times the weight of my 11" Colchester lathe. Relatively simple to setup unless you're doing runs in the tens of thousands of parts where you want all the operations precisely timed so the part time is minimal. These machines can have a number of tools cutting at the same time. Hard machines to make the big money with, a very competitive bidding field to work in where the winning bid on a part might differ from another bidder's price by a fraction of a cent.

[Rich Carlstedt]

Well, I don't do Watch Making , but you can do very small parts on ordinary lathes with the right tooling.
I have seen some amazing stuff made on a Sherline Lathe which is much bigger than a watch makers lathe i believe.
Here is a picture of a 12 inch Clausing Lathe making nuts for a 1 mm thread out of 1/4 rod stock.
The stock was milled hex on a Bridgeport and then presented to the lathe




You can see that normal tooling does not work, so you either buy Jewler's Tools or make your own
It also helps to have a 2- 3 power magnifier unless you have young eyes
Then on the 10 inch Boxford ( ie Southbend) you see a 1 mm screw being made using steel welding rod for stock.
Sharp tools are mandatory and most carbide does not qualify.

Try it, and be not afraid !



Rich

[Paul Alciatore]

In the "old days" they did it with care and skill. And probably a few failures before getting it right.

That small and still accurate? Again, skill and try, try again.

But what is difficult about grinding a 60° tool? We do it all the time. Well, most of us. So it needs that small tip, but if the rest of the tool is 60°, then the tip will also be that same angle. That leaves the amount of rounding or the flat at the tip, which will be rather small. 64 TPI tool would have a flat at the tip that is about 1/10 of the pitch or about 0.0016". That's in the realm of "Oh, just forget it. It will wear down that much before you get to the root diameter anyway." So, just use a fine stone to dress it (polish it) after grinding it on the wheel and you are done. And if you really insist on creating that flat at the tip, then a SINGLE, very light stroke on that fine stone will do it.

By the way, just how accurate does a threading tool really need to be? Lets see, 60°. But has anyone here EVER seen a tolerance on that 60°? I have looked and have never found one. So +/- 1°? Or +/-2°? How about +/-5°? Or even +/-10°? Ok, 10 degrees is a bit absurd. But, seriously, is there any tolerance on that one number of a standard thread specification? If so, WHERE and WHAT? And some of the numbers on threads are controlled with tolerances, like the pitch diameter, but what about the flat and the fill? The tolerances on them is really broad. I spent some time pouring over the specifications for the Unified thread form. What I finally decided was that they were so broad that even severely worn tooling would still produce acceptable threads as long as the pitch diameter was correct for the class of fit. That's just about it. The guys who make the screws and nuts by the hundreds of millions must have had a BIG say in the specifications. A REALLY BIG say.

So just how hard is it to grind a tool to cut those threads? And watch makers probably do not even adhere to the Unified thread form anyway. Anyone ever measure a screw from a watch?

but how did they do it in the old days??? How did they grind a thread tool that small accurately?? Jim

[Stan Stocker]

Originally watch makers used manual processes. Individual watchmakers might have chased the threads, Earliest manufacturers (like pre mid 1800's) made screws using a plate with the same die repeated in many places. The blank rod was turned down to OD for the desired length, the head bottom and sides were formed, and a V cut partway through the blank partly above the screw head to be. The blank was run into one of the empty die plate holes until it bottomed out, at which point when all was done correctly the blank snapped off at the bottom of the V. Once the plate was full of screws to be, the plate went to the screw head finishing area, where they were flattened and slotted. When the V was in the right place, the cut formed a chamfer on perimeter of the screw head, a lead in chamfer on the end of the blank stock in the lathe, and the spot in the middle that snaps off cleanly. Once the slots were all cut and the screw heads were polished the screwed were unscrewed from the plate(s) and put into stock.

Obviously the above and many other watch making procedures are very time and labor intensive tasks, so the American watch companies designed many specialized machines. Their automated screw machines were a huge advance in manufacturing. The threads are cut, not ground. Guys like Edward Howard, Frank Church, and Charles VanDerWoerd were legends in manufacturing circles for their brilliant designs and automated machines and in some cases watch mechanism components as well. At the time, the Swiss were still making pretty mediocre hand fitted watches, and the American watches were the best in the world. The Swiss watch makers all came to the Centennial Exhibition to have a look at these incredible new machines that were letting the upstart Americans eat their lunch.

The history of the early Sewing machine, watch and clock, and firearms manufacturers is just fascinating to me, the breakthroughs, like most others, were huge for their time and set up the basic for modern manufacturing operations. Not just the machines themselves, but also the process management involved. People think Fords Model T assembly line was a new invention, and to buggy and tractor makers it may have been. The folks in the watch and sewing machine businesses probably wondered how else you would do it and compete. Watches had been built assembly like fashion for at least 60 years by the time the Model T came along. Special parts made in specific areas by specifically trained people, flowing in parallel to the sub assembly area, then on to the movement assemblers while cases, dials, bands, and all the stuff that makes a movement into a watch flowed in from other parts of the plant. Shipping at one end of the plant, receiving at the other.

These days, you call the friendly folks at Hermle or Tornos, hand them half a million or more, and they plop a fitted out to your specifications Swiss lathe on your shop floor. Dial it in, load up the bar feeder, and start cranking up the sales. You've got a Swiss lathe to pay off :-) If you want to actually manufacture watches from in house parts you'll likely need a dozen or more, along with quite a few Kern three and five axis mills with automated pallet changers, but hey, what's 750K+ per Kern? We're going to conquer the commodity watch market! Except cell phones have sort of killed that market.

For those still reading and wondering how the die plates were made, the threads on taps were chased and the hardened taps were used to thread the holes lines up on the die plate to be. A jewelers saw provided the cross cuts to expose cutting edges. The final die threads can be polished with a hardware dowel or brass lap coated with oil and diamantine, very fine oilstone powder, or should you have a visitor from the future that works for an abrasives vendor cerium oxide...

Oh yeah, edit here: The die plates were also then used first to make a bunch of taps for the manufacturing part of the operation and to make more die plates. Sort of silly to make all those screws if you don't have threaded holes to put them in.

Cheers,
Stan

[DR]

Well, I don't do Watch Making , but you can do very small parts on ordinary lathes with the right tooling.

.................................................. .................................

Rich

Yeah, maybe so. But, in no way can an ordinary lathe compete with a swiss style machine on parts relatively long in relation to diameter.

[Rich Carlstedt]

Originally posted by Rich Carlstedt --"-Well, I don't do Watch Making , but you can do very small parts on ordinary lathes with the right tooling" ......Rich

Yeah, maybe so. But, in no way can an ordinary lathe compete with a swiss style machine on parts relatively long in relation to diameter.

Stay tuned DR !

I recently submitted an article for publication to our benefactors here ( Home Shop Machinist mag) on that very subject you mentioned
I describe how to do such long/ frail/ small parts with 30 to 40:1 ratios on a ordinary Lathe with a simple tool (not available -you make it ).
Its not anywhere near a Swiss Machine in production rate, but can do some pretty difficult turning for most all machinists........
.....but you will have to wait and see.
Rich