CNC Making Discussion

[Michael_Molnar]

You need to experiment and see what works for your machine and setup. And if you are just hogging out wood, you can move at ludicrous speed. 

[Greg Sigworth]

On 1/26/2020 at 2:06 PM, uncle duke said:

 

I use the drill press method for making my plates - just drill down to certain depths.  The downside is I'm sure I'm the only one doing this in the continental U.S., possibly worldwide - it's a lonely feeling thinking that.

I use the same method. I start with a contour map transferred to the plate then drill down in two or three stages leaving enough space to finish scraping to get to where I want. Currently I am using the Plowden as a model. I can calculate the profile with math on the original contour map. Longitudinal curve is catenary curve and longitudinal curves can be curtate cycloid or currently I am working on circle curves to inflection point. When I did the top I had to adjust the longitudinal curve to have the needed flatness in the center area. I can think the contour in my head and with the correct math make the contour map. 

[uncle duke]

1 hour ago, Greg Sigworth said:

1.   I am working on circle curves to inflection point.

2.  When I did the top I had to adjust the longitudinal curve to have the needed flatness in the center area.

3.  I can think the contour in my head and with the correct math make the contour map. 

1.  You know what gives me fits?  The 6.9 you have to the outer edge is a measurement I haven't been able to nail right the first time.  The depth measurement in between 6.9 and the edge is what I'm talking about - always extra work at both ends of the plate for me.  I believe I had been drilling down to 4.5 or slightly less at the outer edge and living with the consequences later.  My edges for both plates are well under 4 mm when finished. 

2.  I wouldn't worry about the longitudinal arch as long as the bridge area is drilled correctly.

3.  Have you found that a flatter cross grain/arch produces better tone as compared to a tighter, compact radius crossways at the bridge area?  I don't care for the visual of the flatter crossarch but they seem to work better for sound for some reason.

I've seen 6.9 used for Strad contours but I don't recall that being used for a D.G.  Thinking about it one can move the measurements inwards or outwards, if needed.  Never thought of doing that.

My dad says the pain in the hands some of these guys are having during violin making is unbearable.  Hence the cnc discussions.  I just can't imagine what kind of pain yet........  I'll just take his word for it for the time being.

[HoGo]

18 hours ago, Michael_Molnar said:

You need to experiment and see what works for your machine and setup. And if you are just hogging out wood, you can move at ludicrous speed. 

If your machine is too flexible for given load you can also decrease depth of cut and keep the chip load and speed.

[Michael_Molnar]

I do that. Thanks.

[Don Noon]

I figured this might be more appropriate for this thread as opposed to my "bench" thread.

Having a new toy to play with is not very conducive to progress (in the short term) on carving scrolls.  And since I don't like carving scrolls, there's all the more reason to play with the new toy (actually, it's a pretty nice machine).

Here are some taper gages I made out of rosewood, one in 0.1mm steps and another in .05mm steps over a narrower range.  These are for setting the Z axis on the CNC, as well as setting the depth of the drillpress for final graduation marking.

While making the taper gages, I side-milled the edge with a high-helix endmill, and was extremely impressed with how perfect and straight the surface was.  So I got thinking again about using an endmill as a jointer, and idea I have had cooking for a few years.    So I made this one up... a side jointer.   The MDF base has a hardboard lamination, and the maple rail has an ebony lamination.  The depth of cut is fixed at 0.1mm, to be used only as fine finishing.  A tunnel is milled in the rail to get chips to the vacuum system.  It works fantastically well; I'm not sure if it's the shearing action of the high-helix endmill or just that carbide endmills are super-sharp.  And easy to replace if necessary.

Yes, it's way overkill.  But it's fun for a retired engineer.

[ctanzio]

I've used a similar setup but with a router set underneath a stationary table. 

For finishing plate ends, like cabinet doors, or adding decorative edging, I found that the slide rail length on the feed side of the bit only has to be "reasonably flat". But for creating joining surfaces, it does not take much variation along the rail length to get variation in the glue line of the joined surfaces.

What I did was to focus on a "very flat" rail length just before the bit and made sure the wood could be fed snugly riding that small area. I've seen "two point" rails where two small wheels were setup just before the bit and the feedstock always rode along those two wheels. This assured a straight line feed through the bit.

How did you get the ebony lamination "flat enough" to make joining surfaces where the glue line was acceptably thin and consistent?

The other problem I encountered occurred when feeding long stock. As more of the stock traveled past the bit, "chatter" of the stock, lateral and vertical vibration, became an issue. I had to put some sort of "catch rail" setup to block the vibration. Have you run into this problem and how did you solve it?

 

 

 

[Don Noon]

49 minutes ago, ctanzio said:

How did you get the ebony lamination "flat enough" to make joining surfaces where the glue line was acceptably thin and consistent?

The other problem I encountered occurred when feeding long stock. As more of the stock traveled past the bit, "chatter" of the stock, lateral and vertical vibration, became an issue. I had to put some sort of "catch rail" setup to block the vibration. Have you run into this problem and how did you solve it?

The ebony I treated as a slightly thick piece of ribstock, using a sanding fixture on my spindle sander to get it flat and even in thickness.

The maple rail I clamped to the CNC table and took a long side cut with an endmill.  I used polyurethane foaming glue, an accurately flattened aluminum level as a clamping counterform, and clamped the snot out of it.

I drilled a couple of locating pinholes in the base (which you can see in the photo) to mount the rail assembly, and then once it was all together, took a long light cut with the CNC to get the ebony super-accurate with a 0.1mm jog at the cutter location.

I have not encountered any vibration whatsoever in my setup, although the longest stock I will ever put thru it is a viola plate.  There's a lot to be said for rigidity, mass, and extremely accurate collets and endmills when trying to cut stuff.  The spiral endmill I think helps a lot as well, as there's not much in the way of interrupted cuts.  The endmill is 1/2"; I think a smaller one could give vibration problems with sidecutting.

[HoGo]

I cannot find the picture but there was someone in mandolin community who used CNC to join his wedges. He would simply clamp the two halves onto a flat board leaving gap between the halves just a bit smaller than the bit and he would just let the CNC do one continuous pass along the gap milling both sides of joint simultaneously.

Oh, I found the old thread... here:

https://www.mandolincafe.com/forum/threads/28285-Ellis-vacuum-holding-jig?p=354006&viewfull=1#post354006

many interesting ideas even for violin boys in there I guess...

[Don Noon]

I had thought about clamping the wedges to the table like that, but if you're going to join several sets, that's a lot of clamping and setup time.  Much more efficient to mount the jointer fixture once, and then run all the plates thru real fast.  Unless, of course, you add in the time and effort to make the jointer fixture... but for me, that's fun.

[HoGo]

19 hours ago, Don Noon said:

I had thought about clamping the wedges to the table like that, but if you're going to join several sets, that's a lot of clamping and setup time.  Much more efficient to mount the jointer fixture once, and then run all the plates thru real fast.  Unless, of course, you add in the time and effort to make the jointer fixture... but for me, that's fun.

That's not the way real men do it :-)

It's CNC so it's gotta move,no? I've seen folks just use painters tape (two layers held together with CA glue back to back) to hold the plate down. Pretty quick to do. And you only need one spoiler board that can be reused.

[Don Noon]

3 hours ago, HoGo said:

That's not the way real men do it :-)

It's CNC so it's gotta move,no?

Yes, I have to move the spindle to the proper coordinates. 

Just to recap:  this piece of tooling is not justifiable by any rational thought or by any real need.  I just felt like doing it, because it seemed like a fun thing to do.  Old retired engineers get that way sometimes... especially when I have scrolls to carve.

[scordatura]

After spending some time with Fusion 360, I have to say that I am underwhelmed with the loft capabilities of the program. For those not familiar with Fusion 360 that is what you use to create the arching surface. It seems that creating a surface for violin arching gives Fusion 360 fits. I can deal with interface and program quirks but the loft problem is a deal killer. I realize that some have had success but I am finding that I am having to compromise the modeling/design too much or having no success at all despite different approaches. OK the program is free but I am thinking of cutting my losses and using Rhino. After all my time is worth something...I think.

[Don Noon]

I have had good success lofting the central part of the arch (from the lowpoint inward), but that tricky part with the corners is still unresolved at this point.

However, there are other CNC-ish things that may be of interest.  The thing that got me diverted was thinking about counterforms for holding ribs against the inside form.  The ones I made have been adequate, but they were made on my previous machine when it had a slight skew to the X-Y axes, and there were some other things that bothered me as well.  Making full-height counterforms on the CNC was a huge pain and effort,... so I got to thinking about 3D printing.  After a brief bit of research and finding out how inexpensive (~$200) and good they were these days, I ended up with this on my "other" bench:                 And pretty quickly started making stuff...

On the left (in the left photo)  is the first "flexure" design C-bout counterform made on the CNC from stacked plywood.  Too stiff, no flex.  Next to that is a thinner one out of MDF; better, but still too stiff.  Making it thinner would be difficult on the CNC... so the first 3D printed one was the identical design, just to see the difference in flex (still stiff).  Next, I added flex sections.  Acceptable, but while I was doing this I was thinking about improvements, which are in the last (so far) design:  evened out the leg angles, added center spring to put some pressure in the middle, and added a pocket to keep the fence clamp from skating around while tightening.  I had to split the center spring zone to accommodate 7 clothes pins for the lining installation.

I have to say that the 3D printer is extremely nice for not needing setup.  No bits to change, spoilboads to deal with, or how to hold down the wood and the cutout parts, no dust, no noise.  Just push the button, and (many hours later) the part is done.

 

[Chris Llana]

Wow!  I suspect other 3D printing applications will continue to pop up.

[Geigenbauer]

Looking great Don. Thanks for sharing!

[catnip]

Those C counter forms look quite promising.  Pliant structures that bend!

If you are interested here is more:

 

[Don Noon]

Nice video.  Cool stuff.  I designed my first "compliant mechanism" over 28 years ago which was part of the Hubble Space Telescope repair.  Back then we didn't call them compliant mechanisms... just machanisms that used flexures.  Very important for optical stuff, where you don't want backlash or bearings with lubricant.  I like flexures, and use them whenever I can when designing things.

[scordatura]

Wow Don the Hubble mirror problem was quite a bugger. The tolerances are crazy. And I just thought you were a "rover guy".

[Don Noon]

54 minutes ago, scordatura said:

And I just thought you were a "rover guy".

There was life before the rover program. And jobs even before I started at JPL.  I'm that old.

[scordatura]

2 hours ago, Don Noon said:

There was life before the rover program. And jobs even before I started at JPL.  I'm that old.

Don't tell me you worked on Explorer 1. You look very young for your age. Did you know von Braun? Lol

[Don Noon]

I'm not THAT old.

[David Burgess]

2 hours ago, scordatura said:

Don't tell me you worked on Explorer 1. You look very young for your age. Did you know von Braun? Lol

That was my impression too, the first time I met Don. I'm pretty sure that NASA has discovered some secret youth elixir.

[Don Noon]

It's called:  Retiring Early.  I think that watching lots of cartoons helps... at least mentally.

[Michael K.]

On 1/25/2021 at 12:53 AM, Don Noon said:

I have had good success lofting the central part of the arch (from the lowpoint inward), but that tricky part with the corners is still unresolved at this point.

However, there are other CNC-ish things that may be of interest.  The thing that got me diverted was thinking about counterforms for holding ribs against the inside form.  The ones I made have been adequate, but they were made on my previous machine when it had a slight skew to the X-Y axes, and there were some other things that bothered me as well.  Making full-height counterforms on the CNC was a huge pain and effort,... so I got to thinking about 3D printing.  After a brief bit of research and finding out how inexpensive (~$200) and good they were these days, I ended up with this on my "other" bench:                 And pretty quickly started making stuff...

On the left (in the left photo)  is the first "flexure" design C-bout counterform made on the CNC from stacked plywood.  Too stiff, no flex.  Next to that is a thinner one out of MDF; better, but still too stiff.  Making it thinner would be difficult on the CNC... so the first 3D printed one was the identical design, just to see the difference in flex (still stiff).  Next, I added flex sections.  Acceptable, but while I was doing this I was thinking about improvements, which are in the last (so far) design:  evened out the leg angles, added center spring to put some pressure in the middle, and added a pocket to keep the fence clamp from skating around while tightening.  I had to split the center spring zone to accommodate 7 clothes pins for the lining installation.

I have to say that the 3D printer is extremely nice for not needing setup.  No bits to change, spoilboads to deal with, or how to hold down the wood and the cutout parts, no dust, no noise.  Just push the button, and (many hours later) the part is done.

 

Amazing. Could it work for Cello too?