Thanks to Gary Mahony for his piece on “Fancy Gear”
This is the first clock that I can say that I designed “from the ground Up”
I wanted to try a few new features and am really pleased with the result!
The wheels and pinions were designed on Turbo Cad which I purchased for this project and although there is a fairly steep learning curve - there are actually only a few critical program functions that you need to accomplish to make gears. My drawings aren’t as elegant as those that you would get by buying a set of plans off the web but believe it or not the graphic shown here is the complete set of drawings necessary to make this clock.
I wanted to try and build a clock that only had two “arbors”. The tooth designed is taken from Gary Mahony’s web site on tooth designs. It is a cycloid design with all the back lash portion of the tooth removed because as Gary says “clocks don’t go backwards -- ever”.
So thanks to Gary I designed a cyclodial tooth and then remove one half of it and then using that one pattern I applied it to all the wheels and the time train pinions.
Using an Excell spread sheet I was able to check out a whole lot of gearing variations. Although you can’t read them here I was able to test out nearly 100 gearing variations
The objective was to come up with a set number of teeth for all the wheels and all the pinions that would give me a “seconds pendulum” with three identical wheels and identical pinions. The result was as shown in the CAD drawing to the left.
The wheels all have 62 teeth and pinions all have 8 leaves. The mathematical calculation will give a 30 tooth escape wheel a .06 sec/min error on the excel spread sheet calculation. Since I didn’t want a second hand on the clock I knew I could easily compensate for the theoretical error through pendulum bob adjustment.
The net result of all this was that the calculated depthing for all the running gears was 96 mm. With this information I was able to use the cycloid gear design off the web to get the pitch diameter to apply to the CAD drawing.
The next design piece was to design the pallets and escape wheel and for that I choose a Graham Escapement and followed the technique described in “The Modern Clock” by Ward L. Goodrich. The instructions were very complete and by using CAD I was able to get a nice accurate template to cut from that would give a 2º lift on the pallet faces. Once I had the design done I did a straight scaling to achieve a 96 mm centre to centre between pallet and escape wheel.
This is the standard 12 to 1 reducing set which could have been much smaller and finer but I wanted to have everything on two arbors so the 96 mm centers dictated the gearing size shown in the red gears.
The last bit of innovation was in the pendulum. Ward L. Goodrich points out in his book “The Modern Clock” that the only way to have the crutch experience the minimum friction in the pendulum is by having the pendulum pivot at the same center as the pallet pivot point.
In order to accomplish this I opted for a light pendulum and have it pivot on the pallet shaft. In this way the crutch is a linear extension off the top or in this case the bottom of the pallet and straight down to the entry point on the pendulum where there is a 90º bend.
If you are a builder you will undoubtedly point out that since each gear pinion set rotates on the shaft or arbor each effectively has its own arbor and I would agree. The simplicity in building plates is a huge bonus for me since there are no “cumulative errors” with only two arbors.
The upper arbor does turn in its pivots, and that is in order to transmit power to the time train, while the lower pivot (shaft) does not turn.
The total depth between the plates is 3.25 inches.
Assembly naturally takes some time since any spacing with nylon washers must take into account all other items on that arbor, and with stacked gears there is very little room for end play.
The finished clock takes nearly 8 pounds to run and I believe that is because the wheel / pinion sets turn brass tubes on their steel arbors. The objective was not to see how small I could keep the drive weight but rather would a clock run on two arbors with all the wheels in the drive train the same size.
I think that I can say YES!!
I may not win a beauty contest but it sure did answer a lot of questions for me.
And now to the finishing of the clock. I chose to paint the plates black to better show off the gears.
All the gears in the time train were colored a warm yellow. The spokes on the drive train wheels are too thin and I had a couple break during construction. Once installed in the clock they haven’t given any trouble - still next time they will be a little more “robust”.
The unit is now finished except for the weight cover which will be a teak veneered tube that will hide the lead “hockey puck” weights.
Now for the long process of debugging all the little things that cause the clock to stop. I anticipate quite a few with several wheels turning on the same arbor.
I have added a little to the clock
I have added a little to the clock. It seemed to be short of “dial character” so I added a nicer set of hands an the rest of the numbers with only numbers 5 through 7 missing.
The drive weight also got a teak veneered cylinder to make it a little more dressed up.
But the big change was to add a moon dial. The moon dial works on a 29 and 1/2 day cycle. I didn’t realize that the moon cycle varies slightly depending on a lot of things happening up there but a reasonable approximation is 29 and 1/2 days.
In this moon dial the moon stays stationary and slowly waxes and wanes through its cycle. The gear train was “copied” from Jesse Emory except that he used it to drive a calendar wheel. By using a one leaf pinion on the hour wheel and a 59 tooth wheel that drives the “shadow” I get a 29 and 1/2 day rotation.
This was a lot of fun to make - now for the debugging.