The Dale C. Maley Family Web Site

Click here to edit subtitle

Golfing Automaton

I am always looking for ideas on new simple automatons to build.  I searched and found that Wanda Sowry is a very talented lady who builds many different automatons.  I went to her web site, and found a male and female golfer automaton that she designed and built. You can see the female golfer using this link.  The male golfer is basically the same.

Modifying Her Design

I did some experiments yesterday using the Kissing Couple automaton with a 1" throw on the crank, similar to her golf automatons.  Normal speed was 100 RPM and slow speed was 64 RPM. In her model, the golf swing occurs every cam revolution, so the golfer hits air every other rev, because the ball is not returned yet.

I would like to slow the golfing swing down by a factor of 10 or so using a 4 gear set.  The only question is if the swing is fast enough to make the ball go.

Then I want both a male and female golfer on the same automaton.........with the man doing a simpler shot that the woman.  The man might just hit the ball up a hill.......and the woman would curve her shot up a hill.

9:1 Gear Speed Reduction Ratio

If you use 2 gears that have 2" Pitch Diameter and 2 gears with 6" PD, you get a 9:1 speed reduction ratio.  The shafts on the 2" input gear and the 4" output gear can not be connected to each other, for this method to work.

Gear Design for My Golfing Automaton

My Concept Drawing for Combined male and female golfer automaton

Prototype Results

My initial idea to have a cam rub on a disc to rotate the golfer failed.  You really need the upper disc on center with the crankshaft to work best.

So I tried putting a dowel in horizontal disc and a dowel in the cam disc.  This rotates the golfer, but the club speed is way to low to putt the ball very far.........if you are running a low RPM like I want to.

Like a golfer, I need to wind up the body like a spring, then swing and release the stored energy.  This gave me the idea of compressing and releasing a wood spring, made of ash !!  I want to try that next.

Male and female wood golfers

I went on the Internet and found photos of male and females putting. I imported them into Sketchup, then scaled their heighths to 5 inches for this project. I will make them from 3/4" thick stock, the club will be 1/8" brass rod with a wood putter head.

Making male golfer

I scroll sawed out the male golfer.  To get the drill angle right for the club, I clamped him to the workbench, used a paper printout from Sketchup to align the portable drill with a long bit to drill the hole.

Developing Ash Spring Concept

I tried not to use dowels to drive the golfer because the angle is too hard to set.  I came up with a cam from 3/8" thick wood, that uses a regular cam to drive it against the ash spring, then releases it.

Some questions to experiment and find out include...

1. How many degrees of rotation are required to get enough ash spring force to drive the 1/4" ball bearing?

2. Will I need a stop for the club after it hits the ball?

3. Is a 1/4" wide and 1/8" ash spring piece the right cross section?

4. If 60 RPM is crank speed gives about 6 RPM after 10:1 gear reduction, the cycle will be....
       Time 0, start swing backwards of male golfer, then he puts
        Time 3 seconds, man's ball has returned to position on the tee
            Woman strikes the ball, ball returns to tee
        Time 6 seconds, cycle starts over

       This might be too slow, need to experiment with crank speeds.

5. Packaging. Too keep width of toy as small as possible, might need to repackage the putting mechanism, because on the lady,
     it will be on her left side, not on her right side like the man golfer.

Tryout of cam to lift the golf ball

I designed the cam lift with 0 clearance, so I had to sand the parts a little.  The cam profile is about right.  Only problem I foresee is the short dowel and the bottom disc are so light weight, I may have trouble getting them to fall down every time. On other projects, I have increased the wood weight, or added lead filler to a piece to solve this.

1st Trial of ash spring

Hooray, concept seems to work.

As I suspected the cam pushing on the 1/4 x 1/4" short piece on the bottom of the horizontal cam follower, the cam caused the follower to cock upwards some. To reduce this, I added a 1/4 x 1/4" piece on top of the follower, to reduce the amount of cocking.  That helped, but next I will glue on a spacer below the 3/8" thick plywood follower, so the L/D ratio of the hole size to the 1/4" dowel is increased, which should reduce cocking.

2nd Trial of Ash Spring Concept

To reduce the cocking of the plywood cam follower, I glued a 3/4" thick pine spacer below the plywood, then drilled the 1/4" hole thru.   SUCCESS   The spacer fixed the cocking concept, and the ball is consistently.  I made a video of this trial testing. You can use this link to view the video.

When the ash spring is released, the golfer's club rotates past the tee hole, which is fine, but he needs to rotate back to starting position, so this is room for the ball to come up the hole and get on the tee for the next shot. 

The round piece below the hole acts a natural stop for the swing. You have to get lucky once in a while!!

 I need to add another cam to move him back to starting position.

Added cam to return male golfer to home position

Designed a cam with a little shorter throw than the cam that does the putting, and extended the 1/4"x1/4" piece under the cam follower..............and it worked ok.  I had to dowel the male golfer to his 1/4" dowel, because he was rotating each putt a little.  You can see a video of this using this link.

Another option would be to add a weight, string, and a pulley to keep the cam follower against its home position, since the follower rotation is relatively small angle. We will see how it does on the final model.

Adjusting Force of Putting

Just a thought, because I may need this later.   If I made the position of the ash spring adjusable, by moving the holder in and out with a threaded bolt, I could "fine-tune" the amount of force generated for the putt.

Things left to prove out

1. Putting green for both man and woman, and ball return.

2. Putting force

3. Develop desired gear reduction ratio

4. Do I need to shorten model width by repackaging the ash spring mechanism?

Angled Green for male putter

I clamped the green piece against one of my large 90 degree black plastic angles, then band sawed the angle. This worked well.

1st Male golfer putting to his angled green results

I used the dremel with a deburr bit to make the line from the hole and up the angle.  I set the angled piece on the old prototype set-up to try it out.

The initial results were very disappointing, in that only 1 in 10 putts followed the groove and went in the hole !!!!!!!!!  I am not sure what the problem is?

And boy, it is easy to lose the 1/4" ball bearings on the shop floor !!!!!!!!!!!!!


I set up the camera, to see if I could slow it down and see why the ball is not going in the hole.  The first 2 trials failed, and the 3rd went in the hole.  On 1 failure, the ball actually was lifted up and bounced to the right of the straight path.  On the other bad one, the ball went straight, but went past the hole.

I need to try to slow down the swing.  

Another thought, some fails will be ok, just let ball come down and return to the bottom of the slope, then to the starting spot.  Ratio of goods to bad needs to be high though.

More experimenting

To see the impact of a damper, I added a string to the horizontal cam follower, and used about a 3/8" nut for the weight. This reduced the force hitting the ball, in fact it is not high enough now. The string and weight does return the horizontal cam to home position much better than the small wood cam does.  I think I will leave it and remove the small wood cam.

I added a 1/16" spacer to the horizontal cam to increase the ash spring force.  This helped a little, but it changed the starting position of the club towards the ball.  I need to rotate the male body to compensate.

I made the groove deeper for the ball to follow using the smallest diameter drum sander for my Dremel.  But it is time to replace the 3 pieces on the prototype with 1 piece and resume experiments.

2nd Prototype

I decided there was no way I could drill the angled ball return hole on the male green accurately enough with a long drill bit.  So, I scroll sawed the horizontal piece along a line going thru the Tee and green hole, marked out the return passage on each side, then used the Dremel with the smallest drum sander to make the passages.  It took longer than I thought, but I got the ball to return ok.

#2 Prototype

Last try on #2 Prototype

I deepened the guide groove because video showed the ball was bouncing off the RH side of the start of the groove, I also put a slight curve in the putter.  None of these worked well enough. I'm giving up on the idea of hitting the ball and have it take a straight or curved path............I'm going to make a "bowl" and let the ball go around until it finds the hole.  You can use this link to see the 1 success and the many failures I had with this set-up.

Next Design Phase

I am going to make a ball return chute from 3 pieces of wood glued up, versus trying to hand grind a return hole.

Because more lift is required to get the ball up to the Tee, I had to raise the height of the top above the crankshaft, and redesign the cams.

Making the bowl type greens

I made each green with 4 segments.  On each segment, the slope starts at the normal 3/4" board thickness, then tapers to 3/8" at the center.  I clamped a 5-1/2" piece of pine against my 90 degree black square, then slowly fed into the band saw.  I had to use my old trusty Sears radial arm saw to cut the 45 degree angles because my Makita chop saw is out on loan to a friend.

I had an issue of the 4 pieces not being the same thickness exactly along the band saw cut.  I sanded them some on the horizontal belt sander.  I used the wound string and nail method to clamp them up.  I will try sanding them to smooth out the 4 joints on the finished green.

Design Changes

I increased the diameter of the follower for the Tee, to increase its weight, to help it return better each time.

I arbitrarily designed the OD of the dowel that connects to the marble return chute to be 1 inch.  I have some 1" dowel rod in stock, so will probably try to use it.

Ball Return chute

Design as of Morning of March 1, 2021

Building ball return chutes

I sawed a 2-3/4" long piece of 1" dowel, then used my centering jig to drill a 1/4" hole thru the dowel.

Marking radius on chute

I slid the 1" dowel down the 1/4" dowel until it rested on top of the chute, then marked the radius that needed removed with a pencil.  I used a 3/4" sanding drum on the drill press to remove the material. This worked pretty well. 

I could then mark the rectangle where the chute met the center 1" dowel.  I tipped the dowel slightly in the drill press and drilled a 5/16" angled hole in the dowel, so the ball would want to fall down into the 1/4" ID with the Tee.

I used hot melt glue to attach the chutes to the horizontal base, in case I had to remove them to fix them if the ball return did not work properly.

More pieces made for final model

Got the Working Model built :)


Neither ball return chute is 100% effective running 1 ball at a time.  I removed the RH side chute twice and modified it, them hot melt glued it back in place.  But, if you put 4 balls in each side, they run almost 100% effectively !!

You can use this link to watch a video of the model in action.

Desired RPM for golfers

When I made the video above, I tried to crank the model at a consistent speed that was about right in terms of when the golfers each took their putt.  I used the video to determine that the lady made 19 strokes in 40 seconds.  Each stroke is 1 rev.  So, my desired speed is 28 RPM.

From my tests using the Kissing Couple automaton, 100 RPM was normal hand crank speed, and slow was 64 RPM. I'm designing this for kids to operate and they like to crank it up.  So, my gear reduction ratio needs to be 100 RPM divided by 28 RPM, which is 3.57 to one.

From my Excel spreadsheet, a 4 gear set using 2 gears with 3 inch PD and 2 gears with 6 inch PD will give me a 4:1 reduction ratio.

Broke 1/2" wide band saw blade

I had used this blade quite a bit, but don't know how many hours I had on it.  Ordered 3 TPI and 4 TPI Olson blades to try out, from Amazon.

Cutting Gears

It took a while to cut the gears using my scroll saw. I used a #9 blade, and used 3/4" thick red oak.

I increased the distance between the gears from 4.5 to 4.5 inches + 3/32", to give some tolerance for the gears.  The gears assembled up ok in the 3 vertical boards.

Final Assembly

The return chute on the man's side, is still not working very well, even with multiple bearings (balls) in the chute.  I used a pipe clean stuck down the green hole to determine the balls were getting hung up close to the 1/4" vertical hole.

I went into Sketchup and tried to design a more foolproof chute, and them built it.  I used the table saw to get the bottom of the chute horizontal, so the 1" dowel could sit flat on it.  I designed this chute with a 6 degree slope, about the same as the first design.

Broke a ball return dowel

The 1/4" dowel broke attached to the cam follower for the male golfer broke.  It was just a birch dowel.  Will have to replace with an oak dowel.

Tested new ball return chute. When ball went down and  hit dowel, it did not load up on the next stroke, it stayed in place. I used a round file to remove the burrs where the chute hits the1/4" vertical hole. Then it worked with single ball and multiple balls :)   [clamped in the vise]

Had to adjust 1/4 by 1/4 inch block on female golfer

I increased the size from 1/4" wide to 5/16" high, and filed round fillet until it worked ok.

Eliminated 1 set of eye hooks

Drilled vertical hole in weird shaped horizontal rotating piece to put the string through, instead of an eye hook that can pull out, or get in the way.

100% ball return

On the male side, with the new chute, I got it working 100% with multiple balls [bearings] by increasing the stroke very slightly.  I checked my Sketchup drawing, and I allowed 1/4", or in other words, the top of the Tee is 1/4" below the exit chute, but this is not enough apparently. Maybe other stack-ups are messing this up, I wish I had added about another 1/2" of stroke to completely eliminate this problem, but too late now. I dremel drum sanded the small diameter on the cam to increase the stroke, and I belt sanded the follower face to slightly increase stroke.

I can not get the female side to return 100% with the old chute, even though I increased the stroke using the same 2 methods as noted above.  I guess I will build a new return chute for this side also.  Then maybe I will be done building !!

Got the 100% ball return working :)

I made a new chute for the female side, and everything started working. So with 2 balls per green, it is now working at 100% return rate !!!   You can use this link to see it in action.  Note, I left the ash spring actuators off since I am focusing on the ball returns now.


I often put wood parts in my old oven and quickly dry the paint at 120F or so.   I forgot some of these parts are hot melt glued together, and one return chute fell off in the oven!!  Will have to glue it back on again.

Ready to Paint


Uh Oh !!!!!!!!!!

I expected some minor problems with the model running after painting, but it would not run at all !!!!!!!           Both cams were binding up when they started pushing against the ash cam !!   I cleaned the paint from the cam and follower pieces, applied johnson wax, and it still bound up !!!!

Appears the cam is driving the irregular shaped follower up against the bottom of the base, and creating tremendous friction.  

Found the Problem !!!!!!!

It was really irritating to test the unit before painting, then have it not work in a major way after painting !!!!!

After a couple of hours I figured out the problem.  I chose a normal Birch 3/8" dowel for the main crankshaft.  Between pre-paint testing and after painting, the dowel cracked between the male cam and the drive gear !!  You could not easily see the crack until it was under torsional stress.  I replaced the Birch dowel with a 3/8" red oak dowel, and Bingo......everything starting working fine !!!!!!!!

Finished Golfing Automaton

Sketchup Warehouse

I uploaded my model to the warehouse. You can access it using this link.


You can watch this video using this link.

Closing Thoughts on this Project

What started out as a simple idea of having both a male and female golfer, with the man making an easy shot up a hill, and the woman making a much more difficult shot on a curved trajectory, turned out to be much more complicated than I thought !!

I could not get the man to hit a straight shot up a hill. Just as in real life, the ball would leave the club at many different angles. I gave up on this idea.

It also turned out to be difficult to design and build a ball return chute with 100% return accuracy.

The ash springs worked very well for winding up the golfer about 30 degrees of rotation, then quickly releasing the golfer to hit the ball. In the real world, golfers often wind up their body like a spring, then on the downstroke release the energy stored in the spring.

I should have known to use a red oak dowel for the main crankshaft, instead of a weaker Birch dowel.

The grandkids and adults should enjoy this automaton at family and public events.

July 2022 Update

One of my grand-daughters really likes this game, but the gears were getting stuck for her.

Dis-Assembled and found gear set on RH side was not meshing correctly on certain spots on the rotation.  I must have got center distance a hair tight between these gears?  I did several things to make the clearance ok:

1. Put a dowel on a scrap block of wood, clamped the wood to the drill press, and drum sanded the gear OD's to make zero run-out of the gears to their shafts.

2. Used round file to deepen gears slightly.

3. Used flat file to narrow the tops of some of the teeth

I then put both gears on their RH board, pulled to one side on the 3/8" dowel for the large gear, while hand driving the small gear.......identified bad spots and filed. When I re-assembled, both gears worked fine.

Nov 2022 Update

I never got around to putting the golf game back together, because I was busy all summer with outdoors projects.  Now that outdoors projects are done, I put the model back together and got it working ok.  One of my grand-daughter's really likes this game, so she should be happy now.

On the lady golfer, once in a while the upper cam follower binds to the lower cam. I tried sanding a 45 degree angle on the edge of the follower, and put some wax on the cam.  It got better, but not perfect.  The gear set does work perfect now, and it doesn't take much force to cranks the model.

In retrospect, instead of a round and flat cam follower, I should have used a roller type follower?

This is one of the toughest hand-cranked models I have ever built, because of trying to get the golfer's swing motion.

Video of Repaired Model

You can use this link to see the model in action. Note how smooth the 4 gears work now !