CNC and clock building

Anything and everything CNC-Shark-related

Moderators: ddw, al wolford, sbk, Bob, Kayvon

Post Reply
cjablonski
Posts: 227
Joined: Mon Aug 13, 2012 7:16 pm

CNC and clock building

Post by cjablonski »

Based on the number of views and the huge amount of responses I thought it might be a good place to start a new topic, CNC and clock builds. My hope is to create a repository of ideas, tips and tricks utilizing our CNC'S in this wonderful hobby. I know as I progress through my next build, I will try to post pics or video links to its status. To kick things off I will post a couple of clocks I've done since January 1 of this year, both of which are currently in time trials. I truly hope this brings out the amateur horologist in many, and a venue to discuss the hobby. Hope to hear from you!
Attachments
20160211_220337.jpg
1455836420289.jpg
"I'm not smart, I just remain on problems longer"
Albert Einstein

Making many BTU by experimentation. ...some days it gets too warm :)

User avatar
meb
Posts: 100
Joined: Fri Aug 06, 2010 6:49 pm
Location: Nevada, Texas

Re: CNC and clock building

Post by meb »

I wish I had the time to design at least 2 or 3 clocks constructed form wood but right now I can't find the time.

I do hope a few people will be interested enough to join your thread with ideas and knowledge about how to determine gear sizing and weight required to drive the clocks created.

Interesting drive weights on the second clock by the way!

cjablonski
Posts: 227
Joined: Mon Aug 13, 2012 7:16 pm

Re: CNC and clock building

Post by cjablonski »

Lol! The drive weights ( or BB'S) are temporary during time trials and timing adjustment. It makes it a fairly simple task to adjust it by either addition or subtraction of more BB'S to find the right "beat". Some use jugs of water for the same purpose, but accidentally dumping bbs is a lot safer than water :).
Time trials and adjustments are something I do for the first 2-4 weeks after a new child of time is born. They need to find a natural beat through fluctuations in humidity and weather, so I tend not to use the final weight shells until that's done. Once it has around 3 million "ticks" (about 45 days run time) I settle on the final drive shell.
"I'm not smart, I just remain on problems longer"
Albert Einstein

Making many BTU by experimentation. ...some days it gets too warm :)

cjablonski
Posts: 227
Joined: Mon Aug 13, 2012 7:16 pm

Next up- Clayton Boyer's Celestial Mechanical Calendar+ Orr

Post by cjablonski »

So, for anyone who has tapped into this thread, above is what's next. You can find this plan and the plan for his Marble Strike on lisaboyer.com. (not paid as an endorsement. )My plan is to link together his Marble Strike clock, the CMCO and a yet to be designed Westminster Chime. This begins phase 2, the CMCO. As stated in a prior post, I will do what I can to document the process in text and pictures as much as possible. Although this isn't a clock, it's motive works are nearly identical, so the procedure is similar, just more challenging with this than a clock. Basically if you follow through this, a clock is easy :). Plans have been ordered, DXF'S reviewed, material has mostly been aquired just awaiting a few more Internet goodies. The first post will be a week from today, an explanation of what it is along with some progress photos. Hope this is assistance to those wondering how hard it is (really not impossibly difficult ) and how enjoyable it can be :). Hope this inspires some to take the leap on this wonderful field....
"I'm not smart, I just remain on problems longer"
Albert Einstein

Making many BTU by experimentation. ...some days it gets too warm :)

User avatar
bill z
Posts: 342
Joined: Fri Sep 25, 2015 9:09 am
Location: Spring, Texas USA

Re: CNC and clock building

Post by bill z »

I've been wanting to build wood gear clocks for some time. Been working my way there. That is why I bought this Shark.

I've got some PDFs of clocks and this month bought a DXF from Clayton Boyer. Ordered the brass last week. Plan on purchasing the Birch plywood tomorrow (Saturday).

Took the DXF and moved things around in the software so to NOT waste so much wood.

I should start cutting tomorrow after the 'Honey DOs'. From what I understand, it will still take a bunch of sanding and working with the gears to make them work smoothly.

I plan on getting some used wheel weights from a few of the tire stores around and seeing if I can use those and NOT have to buy bird shot or BBs.

I'm excited but the wife just wants me to stop spending money.

cjablonski
Posts: 227
Joined: Mon Aug 13, 2012 7:16 pm

Re: CNC and clock building

Post by cjablonski »

Awesome bill! It certainly is a fun hobby, and once you've obtained the basic materials (brass and ply) the rollover from one to the next gets less and less expensive. Just received my CMCO plans from clayton, and will begin that tomorrow also. Mind if i ask which DXF you are building?
"I'm not smart, I just remain on problems longer"
Albert Einstein

Making many BTU by experimentation. ...some days it gets too warm :)

cjablonski
Posts: 227
Joined: Mon Aug 13, 2012 7:16 pm

Re: CNC and clock building

Post by cjablonski »

PS- if you have any questions regarding the build, ask! I love clock builds, have done a few, and would be more than willing to help!
"I'm not smart, I just remain on problems longer"
Albert Einstein

Making many BTU by experimentation. ...some days it gets too warm :)

User avatar
bill z
Posts: 342
Joined: Fri Sep 25, 2015 9:09 am
Location: Spring, Texas USA

Re: CNC and clock building

Post by bill z »

cjablonski

Yesterday, I ate up 2 square feet of 1/2 birch plywood making what I hoped would be clock parts. Didn't turn out so good even using the DXF from C Boyer. I'm not blaming anyone except me and my newness to this CNC hobby.

The larger gear looks OK but all of the smaller gears are just down right bad.

Knowing just what to do first isn't real clear if you have not ever done this before. So here is what I did and I will be looking for pointers as to what I should of done.

Using the DXF file, I sorted out the parts drawings to be cut from a 1/2 inch plywood. I placed them as being cut from a 24 X 12 piece of plywood in Aspire. After working all of the open and duplicate vectors, I selected all of the holes first and selected drilling tool path using a 1/8 end mill. I had just replaced my waste board and had it marked for the farthest the gantry could travel. So I placed the top of the 24 X 12 plywood at the mark, clamped it down , using the center of the 24 X 12 ply as my starting point, I started the program.

What happened was the CNC was working drilling 1/8 holes but tried to go past the top and the motor started slipping (rattling) trying to go on. From that point on, the holes were off. I stopped the machine and ordered it back to zero and it was some 2.5 inches off. I moved the gantry back and forth to find a better starting point. Something like 2.5 inches down.

My mama told me 'Waste not, want not. I figured to use the same piece of wood the second time.

Again, I had it drill and it did but it just drilled and did NOT make the hole to the size in the drawing. OK, at least I have centers of the wheels and gears.

I had the CNC make the small gears first. They really look bad. Maybe backlash issues in my machine. Bought it used and don't know how many hours were on it. Other work, not as intricate came out good but then again, nothing to compare to. I think the large gear isn't symmetric. On one side the distance between the gear teeth and the back of the gear is thinner than on the other side. So thin you can see light through it in places.

The picture says it all.

I'm guessing that I have the parameters of my 1/8 mill set up all wrong in Aspire, but I don't know what is right. The cuts in the small gears are way to big making the teeth in the gears way to small. Is the 1/8 end mill what I should of been using?

What should the passing depth and step over for a 1/8 end mill be? I have it set to .0625 and .05. I think I took what was set for the 1/4 and divided it in half.
What should the passing depth and step over for a 1/16 end mill be? I set it to 1/2 of the 1/8.

If I need to to something like replace a worn worm gear and nut in the CNC, I should do that before I do more damage to the plywood.

Questions:

What should a 1/8 end mill be set to in Aspire?
What should a 1/16 end mill be set to in Aspire?
What is causing the deformation of the small gears?
In what order should the clock parts be cut out?

Any and all help is welcome.

Yes!! Needing some help. Just not as easy as cut, sand, glue, put together and Ta Da!!
Attachments
Small Gear problem.
Small Gear problem.
First cut
First cut

cjablonski
Posts: 227
Joined: Mon Aug 13, 2012 7:16 pm

Re: CNC and clock building

Post by cjablonski »

Ok my, friend, lets see what we can do here....(funny the timing, I was just working the CMCO in my DXFS exactly when I was notified of your post lol,). Ok, to start....

The Small Gears
This may not be due to the machine being used or the number of hours at all ( Whew, a relief I'm sure). There are more likely scenarios at play here. First, and most likely a contributing factor is the diameter of the bit itself. Although you may have done everything perfect ( chose what you believed was the right tool, set it to cut "outside the line", cut to the correct depth) the sheer small size of the gears doesn't allow enough room for clearance of the bit to not chew through the backside of the teeth. The end result will be gears that appear to look too small ( because they are :). As the tool progresses through the cut, it will not recognize that it is cutting through the backside of the prior tooth, causing it to shrink its size. Basically, "That tool is too big". Found that out myself. Remember, the entire diameter of the tool must be smaller than the smallest diameter at the inside of the tooth, or it won't fit, and shrink the gear. Check the preview image of what the part will look like in Aspire before cutting the material. If it doesn't look the same, the diameter is too big.

How do I make the part then? Well, a couple of ways....
1 way is to purchase a collet adaptor to use smaller bits. Rockler sells a 1/4" to 1/8" collet adaptor which would allow you to use 1/16" end mills. This will provide you the bit diameter you need to fit in between the teeth and not deform the smaller gears.
The Plus-It fits into smaller gears ( pitch diameter is small enough)
The Minus- due to the flute lengths, the bits are only recommended to cut material small than 3/16" thick. Most of claytons gears are 1/2" inch + material. It can be done, but risk breaking bits and tearing up material.
The Second way ( which is how I cut gears mosty) is.........not to use my shark for it.

Surprised? You would think its the best application for it. In most cases, it is. There are plenty of critical parts in a clock I do use my CNC for, just not gears. "Well, then how in H... do you cut them?". Believe it or not....I scroll saw them.....old school. I find that their is a balancing act when deciding what to use the CNC for and what not to. Simply put, I follow this montra...Can I cut it just as well yet faster than programming the shark? If so, old school. If not-Shark. Is it something that absolutely needs to be perfect? Default-shark, but only if I am sure that it is capable ( Example, the frame, pendulum, pallets, escape if I have the bits). If not, old school.

Some tips about drilling, or what I've found successful for clocks. If the hole size is EXACTLY the same size as the bit I am using, I will drill with it. If it isn't I do the following-

The Bit is Too Big( say that's the only bit I have)- I use the shark to "touch off" where the hole is, but don't drill it through. EX- If I need a 1/16" hole but only have a 1/8" bit, I have the shark drill down a tiny amount ( say 1/16" on 1/2" material) and then drill through will my drill press. That way the shark marks the EXACT location of where the hole goes, so I cant go wrong.

The Bit is Too Small( for the hole I want to drill)- I drill through where the hole is going to be, but don't use it to make the final hole size (similar to above). EX- I have a 1/8" bit but need a 1/4" hole-I will have the CNC drill the center location of the 1/4" hole, but use the 1/8" bit. Now that I have the location hole, its off to the drill press to drill the correct size hole.
I have found that the tolerances for the CNC are just too big to make the hole the right size with the wrong bit, and in clockmaking, tolerances for arbor holes is way to critical to have slop.

Some other items you might want to try...
Set your origin (Home location of X,Y, and Z) to the lower left or right corner of your material rather than center. The reason? Convenient. The gantry will come all the way forward for you for bit changes. Additionally, the corner of the material, I find, gives you a perfect place to center your bit. You can "sight down" the X and Y axis of the material to the bit, even jogging the shark just above both axis to see if your material is square to the bit. Lastly, no need to mark the center of your material, the corner brings you home ;)

The stepper motors in the shark are "Dumb". I mean that in a kind way. They have no software to "know" the limits of the machine size. These steppers assume that the machine is infinitely long and infinitely wide. It doesn't know the shark is 24 x 24, so they will continue to travel past that and think " Im still moving", and it will try. That was what you heard when it reached the machines limit, it trying to continue to move, all the while the software thinking obliviously that it is, when you can see it isn't. That is why it was 2 1/2 inches off. Rather than automatically assuming you have done harm to the shark, try this. Use aspire to draw a 2" circle and a 2" square. Create a toolpath that cuts outside the line and run the program on some scrap. Measure the actual size of the parts. If they are close, machine is fine ( and I would be surprised if it wasn't). Only if they are extremely different would I be concerned. I have done the overtravel thing in the past, and haven't had to replace a single part.

Almost done here, 2 more things. Unfortunately I do not use aspire, so I cannot help on that end (sorry). I still use v-carve, which for most things I do is fine ( and fine for clock making). I'm sure you would find tons of help with aspire in the forums, and in particular tonydude is a wiz. Look em up, he can answer those :)

And finally ( thankfully I am sure lol), the order of the build. I would try to follow the plans in the order they were written, but sometimes it makes sense to vary. In general, I cut the frame first. That allows me to have a place to house finished parts and try them as I go along. If something doesn't work right, I catch it right away and can adjust immediately and accordingly, saving me time later. I then cut all the arbors and spacers. What good is having a frame to house parts but no way to hold em in place, eh? lol. Then I proceed with the clock gears themselves, in order they are placed in the frame, back to front. As I get a pair done, I can then check them to see if they mesh well.

That's just my order to the chaos of clock making, your results may vary :D . Trust me, it does become fun. Just keep this in mind. "Your sense of achievement is directly correlated to the height of the mountains climbed....."
"I'm not smart, I just remain on problems longer"
Albert Einstein

Making many BTU by experimentation. ...some days it gets too warm :)

rungemach
Posts: 460
Joined: Mon Aug 02, 2010 8:24 am
Location: Sarasota, Florida

Re: CNC and clock building

Post by rungemach »

Hi Bill

It may also be an issue of going too fast.
The standard feed speeds in vectric software are usually too fast for a shark to use, and keep precision.
That trips up a lot of folks.

I would try slowing way down, and see if your cut accuracy improves. start at 10in/min and work up.
do one gear at a time to begin with, till you get the quality you need.

For clocks I usually just spot the arbor holes and drill then on a drill press to be sure the are at 90 degrees exactly.

Post Reply