DIY CNC Precision Mill Design&Build

[RogerRetro]

Hello Members,

I plan to make various small parts with 2.5D milled features, including plates from aluminum, pcb stock, glass plastics, acrylics, etc. with up to 2000 hole arrays (0.010x.040" pitch) that require stacking, with feature alignment within .0005” These parts are typically sized up to a max of 5x5x0.5”, XYZ.

I would like to make a machine that can perform to these tolerances.

I need travels only about 12x8x4”, XYZ, and I would like a Z clearance of 6" above the table to allow for about 2" for fixturing, etc.

I am presuming something like Xylotex/Mach3 for drive and control, but not set on that. Linear guide types required is not clear to me at this point either.

I have access to a VMC for making accurate frame members, and would like some help in finding a suitable design and electrical/mechanical components.

Looking forward to any ideas, recommendations and advice.

Regards,
Roger

[foam27]

Roger,

I PM'ed you.

[Fixittt]

I too am interested in seeing plans for such a design. I have a whole machine shop at my disposal and a friend willing to make the machines, as he needs something in this range.

[Fixittt]

repost

[wizard]

I've looked at your requirements and have to say that the big problem as I see it would be the spindle. The small hole size will require high RPM to get decent rates. So any machine you buy for retro fit should keep in mind the installation of a high speed spindle.

As to your parts the dimensions aren't that bad so physically you should be able to select fro a number of smaller machines. Tolerances are a little tight for from the factory chinese small mills but you intend to rework the machine anyways. Properly tuned up the machine should be find at the tolerances you want with plain ways. Linear ways of course add benefits, the thing here is that the big pay off here is likely to be on the Z axis. The Z&Y could potentially remain as plain ways, the Z going to ball screws due to the large number of holes you seem to want to drill.

As to the software/hardware combo, I can't say that I have practical experience at this level. It will be a matter of choosing what is right for your needs. I would not dismiss Gecko drives based on their reputation nor would I dismiss other CNC controller software.

Lastly you indicate that you have access to a milling machine. This gives you significant advantage over many in these forums thus the potential of building a machine from the ground up. I would not suggest this though and would go the route of rebuilding Chinese castings as required. That could be a whole mill or just the castiron table off of one. Yeah rework will be required but in the end you get parts that are 90% of the way there.

Based on the cost of a Small Chinese Mill, buying one solely to use as a donor machine can make sense. Just consider the cost of time and materials to come up with a castiron table. Even if your final machine is a gantry system over such a table you have eliminated a good amount of work.

As an aside I've gone over in my head several times a gantry based machine to do things similar to what you describe. I believe this is a good option especially if of a moving table design. Stiffness and precision should be easy to obtain and keep in such a machine.

Dave

[foam27]

Dave...he said .0005"......1/4 to 1/6th of a human hair.....not to be confused with .005"....

Full ball bearings everywhere.....on the linear rails, ball screws, and an air bearing spindle.....if you want to keep those tolerances for any amount of time.

Just my opinion....dosen't mean I'm right or all knowing....

[RogerRetro]

Dave,

Thanks for your input. You make a number of good points. I'm not an engineer, so I need to put my thinking up here for close scrutiny...

Coincidently, I am currently working on a small quick change high speed spindle/motor attachment that I can mount to the Z axis carrier for small hole drilling. I am about to begin experimenting with brushless "outrunner" 10-15k rpm motor/controllers, directly driving a small reclaimed hand spindle. I have a mounting adapter designed and ready to cut for my first trial.

Now, considering I need to mill .5" aluminum, I also need to develop a heavier motor/spindle setup to swap out for the high speed unit. I don't yet really know what is available out there for this light duty milling and could use some ideas in this area.

I have thought about reclaiming an xy table as the basis for a machine, but it seems that it may require additional hight to the structure which would be some compromise to rigidity and accuracy, compared to a router gantry style if the gantry is designed carefuly with appropriate stiffening of vertical supports and use of fully supported guide rails with open ballnuts for travel. I could use some validation/correction on this idea.

At this point, I am particularly interested in the basic design and construction of WidgetMaster's Midi-Router, as seen here on CNCzone, and currently am considering development from that design. I suspect I need slightly heavier and higher precision components than some of what is used there for XYZ drive and guidance.

Along with my spindle endeavor, I am working on a shopping list for suitable component vendors and parts, and will post my results for commentary.

Looking forward to all input...

Thanks again, and Regards,

Roger

[foam27]

Roger,

I am on the SAME path as you. Today I was working on a direct driven brushless spindle as well!

Check this guy out in my research searches...

http://www.cncathome.com/spindles.html

I used to have a Westwind 1HP brushless airbearing spindle with a 1/8" pneumatic collet. It did 0-120,000RPM...yes, 120,000... at 70K RPM you could just BARELY hear it run. ZERO vibration, phenomenal engineering. It was powered by a emerson CT drive, that had it's own interface in windows, so you could set the Accel/Decel...Thermocouple temperature threshholds and torque/speed governers.

It also had recirculating coolant lines going into it, that connected to a Thermotek solid state peltier digital laser chiller. The coolant liquid was an alchohol(sp) /water mix.

The airbearings required 2 CFM at about 50-80PSI. That also required an airdryer and filering system. Sounds like a lot of stuff, but the whole package was not really that big.

I have Pix somewhere if you want me to post them.

Milling 1/2" aluminium MIC-6 at 72,000RPM at VERY high feedrates and low DOC was AWESOME...or shall I say SMOOTH. No deburring necessary.

I should have NEVER sold it.

So, I know what good and professional is...now I too want to make my own. Maybe we can share our progress and ideas. I was thinking inrunner as I only cut at .005" DOC, and would like as high of and RPM as possible to keep the IPM high, and loading low. That results in less distortion, tool wear, and a better finish.

Here is a vid for you of Datron's Velociraptor cutting aluminium with a 60K RPM spindle utilizing a water/ethanol spray mist.

http://video.google.com/videoplay?do...53814431369082

And Datron doing High Speed micro drilling...
http://video.google.com/videoplay?do...12832043105815

[RogerRetro]

After many hours of research, parts procurement and very little building over the last ~7 months, I have a little something to show for my design efforts...

The attached PDF cover page illustrates and summarizes about thirty pages of information that illustrates where I am at on the project.

I am currently working on a spindle drive circuit mod to have spindle reversing for tapping with ER style floating tap holders. The drive electronics should be delivered in some weeks (back ordered dual shaft steppers).

Progress is quite slow, as I have limited access to a VMC to cut & drill the frame material I have, am unsure about make or build of the ballscrew mount bearing assemblies, and am just now arranging to have the arbor cast.

I would be happy to have a critical review, discussion, ideas and suggestions from the group.

Regards,

Roger

[ckm]

Why would you make the z-axis with holes in it? Why not make solid, or at least with a center piece? Also, I would suggest that you are making a mistake by over-engineering the Z. If you really need something rigid in the Z, take a look at fixed gantry, moving table designs....

Chris.

[RogerRetro]

Chris,

Thanks for your input.

I was asked a similar question on another forum, (DIY-CNC@*****.com - New Benchtop Mill Design - ArborDrillMill II) where I replied with the following quote.

"In my mind, it is really quite simple. The form and function of the arbor is to transfer the tooling forces between the Z axis hardware and the Table to the machine's base.

Visualize a truss framed bridge or a lifting crane, common methods of design for strength based on stress transfer through triangulation - simply arranged groups of structural members connected, through which stress and strain is transferred. I apply triangulation principals into the arbor design by just a few points:

1. A wide stance at the base, which transitions triangularly to position the top face where the Z axis hardware attaches. This accounts for lateral (X axis) stresses.

2. The smooth transition of the side profile (arc) that allows for table clearance, and positions the Y & Z location of the Z axis hardware attach point. Also the thickness of the side profile, with a web member between the outside and inside surfaces . These account for vertical & for-aft (Z & Y) stresses.

3. A smooth radius(ed) transition between all members (intersection points). This allows smooth distribution of the stresses between members.

Now, use of an appropriate mass in a cast body in place of somehow connecting straight structural members (beams) provides more optimum stress transfer and resonance (vibration) damping through the arbor.

This is my view, others may not agree."

Beyond that, I believe that a gantry is not as stable as a single upright of a well designed arbor.

Again, others may not agree, and I would like to hear/see some engineering supporting other's claims.

Regards,

Roger

[ckm]

When I was thinking of building my own frame, I looked at a lot of commercial frames to see how they were designed. I suggest you do the same, as there are lot of lessons learned in those designs.

You can look here to start with. It's my collection of frame images. Not saying you are wrong, but if no one else is designs frames this way, there is probably a reason. Those types of curvatures were common in late 19th century machines, but have largely gone by the wayside... It's also quite top-heavy.

Although I'm not a mechanical engineer, I've designed automotive suspension components for a known manufacturer. IMHO, you are not leveraging triangulation. While your structure would be OK if vertical forces were the only thing acting on it, I'm pretty sure it would not work well with the lateral and twisting forces present in a mill. Also, all the open space means that you will have a large variety of resonance frequencies, which is bad for precision milling. Finally, if you are casting this out of cast iron, it will probably crack as cast iron is better in compressive strength and the underside of that design is going to be subjected to large amount of tension.

Oh, and the most precise mills, AFAIK, are actually modified gantry designs like this or this ... Food for thought. And don't forget that mass is absolutely critical, the more the better as it dampens resonances and vibrations which can affect precision and finish quality. There's not much mass in your machine, esp. in the base which is going to flex like mad in that design....

Chris.