Last update: August 2016

Gary's Clock 2003

aka "Junk Yard Cogs"

   Home



Post Script - a couple of things prospective builders should know :

- After completing Gary's Clock, I learned from another builder that there is a fundamental design flaw. The escapement wheel should have 20 teeth, not 18. So, if you decide to build this clock, you will have to design a new wheel.

- I was also advised that the D.P. of 10.35 on two of the gear was wrong.

- Gary emailed me the following : " I think the main issue with the clock was I was too efficient in adding the little push the pendulum needs to keep running.As the weights fall they are stopped by the escapement mechanism.The entire gear train comes to a complete stop and starts up again when the pendulum retraces its path. The way the escape part is shaped ends up giving the pendulum a very small push to keep it running. If you didn't do this the pendulum would eventually stop due to friction and wind resistance.


In the 2003 clock I placed the escapement near the bottom of the pendulum. My thinking was this was more efficient. Turns outI was right. It is more efficient. But too efficient.


The inertia of the gears and the placement of the escapement have to be balanced to work right. It's all about balance.
I've noticed that quite a few people have tried making the clock from aluminium and even brass.This changes the inertia of the clock a lot. They managed to get it to work. "

- And finally, Gary advised that he never made the clock. That in itself if not necessarily a bad thing as it should not affect the design which I  find is beautiful and should not stop anyone from making one for themselves. But I think it is a material issue and prospective builders should know this.

I have recently (2016) built two of Clayton Boyer's clocks. The first one, Simplicity was a great learning exercise and runs ok. If I  built it again, now being a bit more experienced, I'm sure that it would run like a top. The second one, Solaris, runs beautifully and is gorgeous. Whether you  cut your parts manually or with CNC, you will spend many hours of your own time in a clock project. While the appeal of free plans of Gary's clock may be appealling to some, I very strongly suggest that you pay a few dollars to Clayton Boyer, Brian Law or Wooden Times. They all make beautiful designs, their designs run and are well tested, and there are many people who have built clocks from their plans. The designers are very engaged with builders to assist when necessary. The broad base of builders of these clocks give the new builder a vast collection of experience ready to be passed along. Finally, go to Youtube and search on these names to see many running examples.


This is my first real CNC project so a few parts have been cut and then re-cut. Here's the results so far...

"Junk Yard Cogs" refers to my shopping habits. Almost 100% of the parts below are made from junk yard scrap.

I found these plans on the internet. It's a skeleton clock and was actually designed to be made of wood. Everything so far is brass and aluminum. The thick parts being aluminum and the thin stuff in brass. I hope my efforts will do justice to this beautiful design.

Making this clock takes a huge amount of time. For example, a 3'" x .250" gear takes about 5 hours to make. Between fine tuning the DXF drawing and setting the cut path in the CAM software, and then the actual cutting and finishing, it's a very long process. Starting this from scratch, without a drawing or CAD file would be a career in itself I'm sure.

Gary deserves a great deal of thanks for posting his plans in DXF format without charge. The satisfaction in completing each part is absolutely immeasurable. THANK YOU Gary.

Gary's site seems to be missing in action. He gave his plans aware as a free download. Since I got mine free, and since I've been asked by others for copies, I will post them here for download just as Gary did.

 

Plans no longer available from this site. 

Unauthorized Photo courtesy of Gary, the clock master. 


This gear is about 2" in diameter. The largest one will be a bit over 7" which is just within the limits of my Sherline CNC mill.

Some of the finished gears. The Loonie is in the picture to provide a sense of scale. For you non-Canadians, a Loonie is about 1.125"

The extra two holes on either side of the centre hoe in all of these pieces is so I can mount the part in a mandrel and mount them in my lathe chuck and also so I can attach the gears to the shafts with machine screws instead of glue as called for in the plans.


More of the brass and a $US which helps some of you get a better sense of size. Only one piece has been tumbled and polished so far. 


This is the base to which all of the shafts, gears, etc will be attached. After proving to myself that I could indeed cut some of the more intricate parts, I thought it was time to make the base so I'd have a place to attach things and to test those parts for fit and accuracy.

The base is actually a refugee from a scrap yard. I found a piece of 1/2" aluminum plate a bit over 30" wide. Since the plate is 29" long and way too big for my little Sherline mill, I had the local water jet cutter whip it into shape for me (and a spare just in case). He also pierced the hole locations. $60 and worth every penny.

You can't really tell from the picture, but it's been carefully wet-sanded. And, I beveled the edge with a 15 degree router bit. I may get another bit and push it closer to 30 degrees. I'm thinking of engine turning the face but it may detract from the rest of the mechanical parts which are really the start of the show. The brushed look from sanding looks quite nice.


With great care and a quick practice run, I milled the first hole recess. Good thing to as I was not properly setup for the first one and all the work so far would have been for naught if I blindly let the mill do it's thing. This will be the most time consuming piece as any mis-steps on the face of it will turn it back into scrap metal. ( I do have a contingency plan in mind though - larger diameter shaft bases.)

The parts are starting to be fitted to the base. Now it's easier to figure out what needs a bit of re-engineering.
The original plans are based on wood construction. As such, some metal and plastic parts are called for 
because of the movement of the various gears and shafts. Now that I'm better understanding it all, I'm having
to backtrack a bit. With the slow movement of the gears, I should be able to easily change a few things as all of my parts
are already metal. The plan is to replace some nylon bushings with brass. And the end supports are 
being replaced with two brass columns at each end. You can see two in place on the right. The two shiny stubby
looking stops over by the gears look like chrome but they're aluminum which I polished with Tripoli and a white polish. 
Both the aluminum and brass can be polished to a mirror finish; the question is how long they'll stay that way.


P19 wheel
about 4" in diameter. 

I've changed this a bit. The plan called for a different decorative hole pattern ( at least I hope it was just decorative). The original holes came too close to the center hole for my needs. I needed to drill two additional holes - one on either side of center. These are 1/8" each. See the next photo.


P19 wheel above
 and
P20 10 tooth pinion.
 
The latter is a bit under 1" in diameter.

The plan called for the 40 tooth wheel and the 10 tooth pinion to be glued together. I didn't want to do that. So, I drilled two additional holes, 1/8" diameter, 1/4" on either side of the center in each gear. Then, I inserted a piece of 1/8" aluminum rod into each and trimmed to length. They don't need to be glued or screwed into place as they get sandwiched between the back side of the shaft and the washer and socket screw on the front. This locks them together quite nicely. See next photo.


Instead of a tapered pin to hold it all together, I cheated. I made a brass washer which was actually one of the brass pieces cut out of one of the holes on the large wheel shown here. I then drilled and tapped the shaft that the gears spin on and turned it to length so that when the socket head screw is tightened, the wheels still run free but with little or no slop to them. I'll be getting some nice stainless steel button head socket screws to replace the one show here. That should help to finish things off with a little dressier finish.


I'm still experimenting with the attaching hardware. You can see the S/S button head socket machine screw
 in the left pulley and in the brass bar (still unfinished). 

The brass washer is one I made for another part of the clock; it's just here to check for appearances and fit. 
The brass looses its shine very quickly. A clear coat will be in order. In the right side pulley, is a brass machine screw.


The weight pulleys.
The one one the left is nearly done. It still needs a bit of cleanup inside the holes. The right side is fresh from the milling machine. Each pulley takes about 4 hours to mill and then another 2 to 3 hours on the lathe cutting out excess on the spokes to lower their profile, cutting the groove for the line and then a little more time with the file, some sand paper, and polish.

P36 Pulleys for weights (4.50" diameter)



P31 3.125" Ratchet Wheel


P33 Ratchet Striker (3.028" long) 
and
P34 Ratchet Striker Spring (4.506" long)
Parts for the winding mechanism


P43 Pinion (1") and P48 wheel (3.2")
A number of drawings call for the wheels and pinions to be glued to the shafts. Rather than do that,
I drilled and tapped the pinions and drilled the wheels for #4 - 40 machine screws. I made a brass
spacer of the required length for proper spacing. Initially, I used two screws but after thinking about
it, I thought that the pinion and wheel may not stay in alignment. Since my geometry teach told me
any three point form a plane, that's what I did. Very little extra work involved and it's all stronger.


P22 Pinion and P23 Wheel
I always find that it's so difficult to show the richness of the highly polished aluminum and brass


Progress to August 20,2006
Some parts will have to be made again as my changes in design continue to change. Mostly shafts & spacers ; easy stuff.


P53 Pendulum Top Disc
I started with a piece of .500" aluminum and then milled down from there. 
I drilled two holes for each pendulum stick. 
A piece of .375" brass will act as both a mount to the base and as a bushing.


P53 Pendulum Top Disc
Completed with mounting hardware


P57 Pendulum Weight - in progress
I didn't have a block of steel for the weight but I did have some 1.00" aluminum
Since aluminum is so light, I milled out two pockets and will fill them with melted lead. 
My arithmetic tells me that I should be within .010 pounds of the weight of the steel recommended by Gary.

CNC is faster than doing the machining manually but this little 1.625" x 1.75" x 0.750" pocket took 1.5 hours
Depth of cut was .020" and feed 5.20" per minute


 

P57 Pendulum Weight
Finished except filling with lead for prescribed weight of about 6 Lbs.

P54a Lower Pendulum Disc
This is part "a" since I have redesigned the single part wood disc into two parts.
This is the back disc and is aluminum. I have my eye on some copper which will be a bit smaller
than the diameter of the inner circle engraved on this disc. They will be spaced about .250" apart
to provide room for the weight retaining screws to slide up and down the two slots.


P54a Lower Pendulum Disc - The back side view
This now has the weight and weight adjuster block installed. These will both be drilled vertically
and a piece of stainless steel threaded rod connecting them topped off with some sort of knob / handle


P54a Lower Pendulum Disc - The front view
Here you get an idea of how things go together.


P54b Lower Pendulum Disc

This is the brass face for the pendulum disc. It's made from
 .080" brass which I engraved with a simple circular pattern.

It's still fresh out of the mill and needs some wet sanding
and maybe a bit of polishing. A bit of clear coat will also
likely be in its future. Likely clear power coat.

 


The Pendulum Assembly
Overall length - 41 inches

Left: Quickly assembled to test fit... which was perfect !
The aluminum vertical strips will be trimmed to match
the contours of the top and bottom discs.
(The blue stuff is just layout dye.)

Right: Final assembly


P32 72 tooth wheel in aluminum -  in progress

P32 finished. About 7.5" diameter. 
This is the biggest wheel of the project.


A12 Drawing Assembly
P32 in Brass. After all that effort making it out of aluminum, I decided that it was
 the center piece wheel on the clock due to it's size and really needed to be made in brass. 
Also, with the P33 & P34 aluminum ratchet parts in aluminum, the brass provided
 much better contrast to show them off. Holding it all together are the P29 shaft & P30 drum.
Don't you just love the fit that CNC provides for the ratchet.


This is what happens when you make this clock from metal and not wood. All the connections need to be mechanical, not glued.
This is the P30 Ratchet Shaft being drilled to connect it to the P31 Ratchet Wheel

Fully assembled and ready to mount on the base


P 26 Cannon Tube tapping with #4 tap
Cannon tube attached to P 27 Wheel on right


P 28 Hour Hand attached to Cannon Tube

Every now & then something doesn't go quite right. I drilled a .375" shaft hole a wee bit off center. The only solution seemed to be to open up the hole and put in a separate piece to support the shaft. But it had to look like I did it on purpose. Which of course... I did.

So the .375" hole was opened up to .500" and two #10 machine screw holes were drilled on either side.


The slots allow for some movement to align the shaft.

On the right, you can see it earning it pay.



 

 

 

Updated: September 07, 2016