#5 Regulator progress

Next jewel installment: The Levin compound has a receiver for several attachments. The hole grinding and polishing jig I made myself, the little grinding head came from somewhere, and I adapted it to the receiver.


You will notice that the jewel is already polished flat on top and round on the sides; it is the same process as the rough grinding in the last post but using 1200 paste on different lap. First I bring the hole out to nearly the pivot diameter, checking with gauge pins. This goes pretty quickly, and I have changed to 1200 paste in the second photo. It will still increase the hole size as it smooths out the inner surface. I clean and check the hole, and add paste many, many times in this process. The polishing head is swung around so that the acting surface of the tapered lap is parallel with the hole. The working surface should rotate against the direction that the inside of the hole is turning. The small laps are made from copper wire, and are threaded into the end of the shaft. I have assorted sizes, ranging from #8 copper ground wire through #14 copper house wire. These are used for polishing lots of things.


Next I grind the oil sink and smooth it with 325 paste on one of the larger copper laps. Later I'll polish it with a shaped piece of pegwood charged with 1200 and then 14000


After the process of smoothing the sides and face, I'll chamfer the end I'm working on. This actually occurred prior to hole grinding and polishing, just after the second polishing of the sides and face of the jewel (not documented), and while the polishing head was still in the compound.


At this point I'm ready to make jewel chucks for holding the jewel to polish the other end, and to final size and polish the hole.
That will come next.
Johnny

 

This concludes the jewel-making process. I have to polish the other face of the jewels, as well as the other ends of the holes. Each chuck is made to fit both a 5C and a WW collet, for purposes of sawing the slots, and the final polishes on the jewel. Two of them are within a couple of thou of being the same diameter, allowing me to use one chuck for both, which is why there are only three.




Making these and these and the pallets is a huge job. If anyone would think this is an arcane and insane endeavor, they would be right. When all is done, they are pretty satisfying, though. Next I'll make the bezels and mount them in the plates.
Johnny

 

It's been a while. Too much skiing, traveling, and recently whitewater rafting. I finished the bezels some time ago, however. The bodies are fit to each jewel, which is set into place, and the lip of the tiny flange is burnished into place. The large O.D. is made to fit the #80 collet, and the small O.D. is made to fit #65, for convenience in shaping and burnishing. The bores are made in the plates, and the tiny tap holes are drilled. I don't have pics of the countersink process, but they are made with a small screw-head countersink burr with a pilot, before tapping #0-80

After making a bunch more tiny screws, and several larger ones, the bezels are mounted, and the plates begin to look ready for motion work on the front, and the movement and pendulum support brackets, along with a back plate.

More soon,
Johnny

 

Now the motion work. This little train is 15t-45t, and 12t-48t, for a reduction of 1-12. The 15t cannon pinion is turned, drilled, and then the teeth cut. The cock is roughed out, the wheels cut, and the pinion repurposed from some old mistake. Crossing out and polishing is as before. This completes a huge part of the work; it runs quite well on the maintaining power. I'll leave final polishing of everything until I get brackets, mount plate, and pendulum made.
Johnny

 

Thank you Harry - It feels like I'm "over the hump".
Johnny

 

Thank you Phil. The cannon pinion will have a small "canoe" spring under it, pressing on the outer shoulder of the pivot. This will be held in compression by the hand collet and pin. A more robust arrangement is for this spring to be mounted on a square at the outer edge of the pivot, but I don't have the room for one as the pivot and shaft are fairly small, and will make do with seating it on the edge of the pivot.
Johnny

 

Phil - after considering your question, I measured the length of my center wheel front pivot, and realize that I may have enough length and diameter difference (between pivot and minute shaft) to form a partial square on the end of the pivot. In my previous clocks I think I had made the pivot too short, but in this one it's long enough. If I can make it fit, it's a much better arrangement. Thank you for reminding me of that.
Johnny

 

I'm getting to the nitty-gritty of mounting the frame brackets and pendulum suspension block to a back-plate. This relationship is pretty important, as the pallet arbor axis must be as near as possible co-axial with the a"center of rotation" of the suspension spring. We all know that this changes somewhat with the arc (contributing to more or less circular error), and changes the timekeeping. What I can't find anywhere is a reasonable approximation of the distance below the suspension spring clamp for the center of the pallet arbor. Frodsham, Goodrich, and Rawlings all say something like "Make it the distance that works best for you". I have always used something on the order of .050" below the clamp for a fairly heavy pendulum (say ~12 lbs) with a .006 x 1/2" spring about 3/4" long. Has anyone ever seen a good source of information about this?

More pics coming soon
Johnny

 

Kilgore - This would work IF the pendulum suspended on a spring swung in a circular arc. However, as the pendulum arc increases, the "virtual point of suspension" changes. As the spring deflects more, the curvature at the top end increases, and the thus lowers the point of "virtual suspension", and decreases the radius of arc. This much is just due to the increasing resistance of the spring to deflection. Additionally, centrifugal force on the mass of the pendulum bob tends toward sideways as the arc increases. This is what causes tall cases to sway and absorb energy. This force effectively increases the weight pulling on the spring, and gives it a sideways vector near the ends of the arc, confusing the visualization of all this. Others have spent time and words on the problem, but I don't know of any definitive analysis. This effect has also at times been seen as an incidental correction for circular error of the pendulum.

Any of our resident engineers are welcome to chime in.

I have used a distance from the point of the spring clamp of .040" as my best guess. This is for a fairly small pendulum arc, and may be off somewhat.

Thank for your thoughts,
Johnny

 

It occurs to me that the "sideways" force on the pendulum suspension near the ends of the arc may be more the effect of that component of the gravitational return force on the pendulum as it nears the end of its swing. Maybe I'm crazy, thinking about this.
Johnny

 

Some progress in spite of summer stuff: Backplate and movement brackets, to mount it all in a case.

The plate notches and bracket stops to keep the movement exactly in the same place after removal. The brackets are dowel-pinned to the plate, so that they can't ever be moved.


The pendulum hanging block, also dowel-pinned to the plate. In my other clocks there has been some adjustment for the pendulum height and location of the movement, which could eventually lead to running problems. This was why the head-scratching about the location of the pendulum center of rotation.

The pendulum is in process, and I will have pics soon.
Johnny

 

You are very welcome guys. More soon,
Johnny

 

More minor progress: drilling and slitting for the suspension spring in the rod end of the suspension assembly. This gets mashed down a tiny bit to tightly fit the suspension spring.


The steel components of the pendulum stirrup, with a dummy rod (no adjustment threads) in it.


The new rod end and threads, being milled to fit into the stirrup top.

 

Thank you for the reference Phil. That book is currently unavailable, but I'll keep an eye out for it. In my other clocks I also used slotted adjustments for the positioning. I decided in this one to make it correct and fixed, so that down the decades and centuries it couldn't be changed for the worse. The pointed rod seems like a good technique. I looked through the front jewel bezel opening with a binocular microscope to align the rear jewel pivot hole with the bottom of the top spring chop, and then lowered the movement about .040", as that seem "about right".
Johnny

 

Thanks Tok - I'll try that. They were quite responsive about a shipping problem with the recent Betts book. I've long admired Woodward's approach to precision clocks.
Johnny

 

A bit more progress:


These are things that are not trivial, taking a bit of fiddling around to get right toward the end, but have to be completed. The clock is actually now running, though without dial, hands, or the mercury jar. I just have a slug of metal in the pendulum stirrup for test purposes. I'll get some pics up of it soon, but am concentrating on the dial at this point.
Johnny

 

Thank you Paul - I surely will show the final steps, and you're very welcome. I'm working with a cabinetmaker for the case - I won't have as much to do with this one.
Johnny

 

They are heat blued in various trays filled with brass filings which conduct the heat evenly. That's a traditional method used for hands, as well. Another part of the trick is getting a high polish and making sure the surfaces to be blued are scrupulously clean. The rest is long practice including do-overs.
Johnny

 

Some recent work: Indexing the rating nut. With 36 TPI (#8-36, in this case) and 30 index marks, each mark will change the rate by close to one sec. per day.


Dial work on the little Dyna mill. This is where it really shines. Others do this by photo-etching or somehow else, but this is really clean. The programming is not trivial, and being limited to 10k rpm, I have to move pretty slowly with the .020" end mill. This took 7 hours, and had to be done in 8 segments, due to the limited size. However, it's done, and a heck of a lot more easily than the several I've done on rotary tables. I'm working up a mount for an air spindle (50k to 80k rpm), for future engraving and such. Cutting the disc out of the octagon was done on the rotary table on the big mill.



Hands, bezel, silvering, and finishing details next.
Johnny

 

Thank you Paul - I have to engrave a "Name and Number" on the dial with a pantograph, and then wax and silver it. I'm kind of anxious to be done with this one. We have a cabinetmaker making a case right now, so it shouldn't be long.
Johnny

 

The bezel starts out as rough work:

After silver soldering it, I mount it on a piece of MDF on the rotary table, and proceed to mill the inside true, and form the outer and inner flanges.


Finally I drill and tap the holes that will both hold it on the lathe faceplate and mount the dial.



More to come, I'm getting close.
Johnny

 

After the bezel is screwed to a hub on a board on the faceplate, I turn the outside and front face true. Then, with the top slide turned around to about 70 degrees, I make the inner bevel, moving the carriage .005" at a time, using the adjustable carriage stop. The dial is now engraved with name and number, and the hands are ready to mount on their hubs and blue.
Johnny

 

Thank you Allan - that's 1/2" x 3/4", bent the "hard" way. The Hosfeld style bender actually makes it not *too* difficult. I call it my Bendito..... The setup on the rotary table was the time-consuming one - I had to "hopscotch" the clamps around the table so that the piece never actually came loose, and made quite a few passes.
Johnny