General Discussion

Clock 13 Tuning notes

As noted in my 'show and tell' post, I did a hasty initial assembly and my clock ran for about 2 days on 4lb 6oz direct weight. Then it stopped and refused to restart.

My original bearings had metal seals that seemed impossible to remove. I got some of the rubber-seal type and removed the seals and lube per the instructions. The lube was a translucent stuff reminiscent of petroleum jelly. IPA didn't seem to dissolve it, so I used acetone.

I found that the dry graphite in the bearings seemed to make them worse, though I suppose it might improve if I ran it for a while to grind up the granules. This was one of the large skate bearings. I skipped it for the rest, and am running them dry. They appear to be chrome plated, so I don't expect corrosion to be much of an issue.

I got out my loupe and carefully went over all my gears, scraping away any blobs or stray hairs that were still present. Oddly, I found this was easier to do with a fingernail than an Xacto knife. Under magnification, I could see that the knife was making as many new hairs as it removed. The fingernail cleanly broke the blobs and zits free from the surface. Not advice; just my experience.

I assembled the frame with just the pendulum to test for free swinging.

These are my results:

minutes test1 test2 test3 test4 test5 test6

0 5° (est) 5° 6° (est) 5° 5° 4°

1 n/a 4.2° 5° 4.2° 4.3° 3.5°

2 n/a 3.4° 4° 3.8° 3.8° 3.1°

3 4?° 2.8° 3.2° 3.2° (est) 3.2° 2.7°

4 2.5° 2° 2.6° 2.8° 3.0° 2.4°

5 2° 1.6° 2° 2.6° 2.7° 2.2°

6 1.8° 1° 1.4° 2.2° 2.4° 1.9° (est)

7 1.4° 0.6° 0.8° 1.9° 2.2° 1.7°

8 1 .2 0.4° 1.7° 1.9° (est) 1.6° (est)

9 0.6 n/a ≈0 1.6° 1.7° 1.4°

10 0.2 n/a n/a 1.4° 1.6° 1.2°

11 n/a n/a n/a 1.2° (est) 1.4° 1.0°

12 n/a n/a n/a 1.1° 1.3° 0.9°

13 n/a n/a n/a 1.0° 1.2° (est) 0.8°

14 n/a n/a n/a 0.9° 1.1° 0.7° (est)

15 n/a n/a n/a 0.8° 1.0° 0.6°

16 n/a n/a n/a 0.7° 0.9° 0.6°

(est) means my attention lapsed slightly when the minute turned over.

The first three tests seemed to be getting worse with each test. Then I noticed that the front shaft spacer was missing, and I hadn't tightened the rear setscrew...so maybe the assembly shifted axially and was rubbing the axle end on the frame.

I reassembled more carefully, then restarted w/test 4. The results speak for themselves.

I then assembled without the escapement. I rubbed dry graphite lube on the arbors while assembling. I got a proper bottle of lube this time, but in the past I've had good luck by simply applying graphite from a pencil. I let the whole train spin freely until my test weight hit the ground twice. All seemed good, with no points that seemed to slow down or run unevenly.

I reassembled with the escapement, and started her up, with weight reduced to 2lb 1oz (direct weight; no pulley yet), which is less than half what I needed before. I get almost exactly 2° swing each way. The clock has been running this way for over an hour as I write this; will report here with any problems. I'm guessing I could go down even more on the weight.

I now have a theory about why the clock stopped after the first two days. I think the spring in the friction clutch on the center arbor (minute hand) overpowered the set screws, pushing the pinion forward until it contacted the hour hand wheel, thus causing excess friction. I noticed this only on reassembly, so I can't be sure this was the culprit; I might have bumped it out of whack while assembling.

I used a heavier spring here than the plans call for, simply because I have a bunch of springs to choose from and felt like it made a more solid connection, but if all's working as designed there should be no need. If it stops again, I'll check for this first. If it proves to be the case, I'll switch to a lighter spring.

Note that I had to reprint the set screw-collar piece (the rearmost part on the minute-hand arbor), since the first one split when the screw was driven in. I think this is a weak piece, and ought to be enlarged a bit. Before driving the M3 screw on the replacement, I drilled out with a 7/64" bit (which is ≈2.78mm, so a good size for M3). Another option would be to simply use a steel shaft collar here. Ten of these can be had for about $6-$7.

I consider myself a fairly accomplished designer, but learned an awful lot from studying this design and Steve's design guide. Kudos! I'm anxious to try out the 'perfect print' gears, and do more testing with cycloidal gears in general. I already have a Fusion 360 macro in the works.

-Matt