DIY CNC Mills

[i_eddy_arv]

Just to throw a question out there,
I'm building a CNC Mill based on the mantis design,
has anyone made one of these and are there any pitfalls I need to consider before cutting the material and building the mill

Specifically the spindle and endstops on the X,Y & Z Axis, 

the plan is to use the mantis mechanics, adapted for metric parts, then the reprap gen 7 electronics,

Ideas welcome!

I.

[pickit2]

Pitfalls, Its made of plywood, that's one big pitfall.
Good point is its made to be easy adapted to your use.
Bad point are there is a lot of "lets glue this on to that"
The ones I have seen don't have to much support for the gantry, and bushing for the rails bonded to the ply, so no ajustment, can be made.
A good spindle will loosen all that bonding.
Then there is the most important question, What are you going to do with it, save it for winter fuel?

As its a moving table design I would invest in some good rails and a screw, just build the X axis and get the table moving, then add the Y axis.
If you get this far, then get the Z axis fitted.

[i_eddy_arv]

Thanks for replying Pickit2

I've made some mods to the mantis, its made of 6mm acrylic, and no glue,
all of the bushings are mounted on laser cut acrylic housings, not epoxy'ed onto a plywood member, and all of the motors are also mounted on screwed on acrylic platforms.

I'm using a laser cutter to get good clean edges and hopefully precise holes for mounting the rods,
hopefully ruling out the need for matched drilling.

Definitely wont be burning it ( imagine the fumes )

The proposed use for this is to mill PCBs with 0.5mm track and gap clearances.

[Ichan]

...reprap gen 7 electronics,

Seems it is not powerful enough for a milling machine, what motor?

...its made of 6mm acrylic, and no glue,

6 mm acrylic still too easy to break, need at least 10mm imo.

-ichan

[i_eddy_arv]

Hi Ichan

How would I go about doing the sums to see what depth of acrylic I should use?
I'm an electronics eng not mechanical.   

Its a NEMA 17 47nm torque from ebay,  no brand

i.

[pickit2]

acrylic has a problem with stress cracking.
http://www.directplastics.co.uk/acrylic-sheet

[CocaCola]

acrylic has a problem with stress cracking.
http://www.directplastics.co.uk/acrylic-sheet

Yeah, most of the cheaper 'plastic' machines use HDPE or sometimes Delrin/nylon depending on cost...  Of course gluing HDPE can be a challenge if the design calls for that, but it can be welded fairly easy, or mechanical fasteners used for wood work well...

[i_eddy_arv]

Thank you all for your advice

This has got me thinking on two lines of thought:

Plan A will be to go with a 6-10 mm acrylic machine ( the laser cutter will tell me which )
and a basic circuit to test the stepper motors, then validate motion on all 3 axis, after that I'll look at a spindle assembly,
could even do a permanent marker to just write on copper then etch ( the goal is to make a PCB mill)

Plan B a RepRap Prusa/mendel might also be a candidate for the job?
anyone ever hear of a dual purpose reprap,   3dprinter by day  PCB milling machine by night?

RepRap looks like they are capable of the job but I dont see to many of them working as milling machines.

I.

[CocaCola]

could even do a permanent marker to just write on copper then etch ( the goal is to make a PCB mill)

Ironically this is actually harder then milling out the traces on DIY and/or cheap mills, as the bed and PC board have to be dead on true to the spindle or else the marker won't write consistently across the entire travel...

[pickit2]

If I had some free time, I would build this.
http://www.diyouware.com/front

[CocaCola]

If I had some free time, I would build this.
http://www.diyouware.com/front

That is a nifty design, looks like it could be a lot of fun with decent DIY results...

[i_eddy_arv]

Thats Really good, 

Still have to etch the PCB, but it would probably make very fine tracks thou.

The unstated goal of the project is to remove dangerous chemicals from the PCB manufacture chain.




Posted on: September 16, 2014, 02:28:27 14:28 - Automerged
Progress So Far,

going to finish the design then decide to move to a different platform entirely
(was a great way to figure out Solid works )


Future Work
I'm leaning towards a dual purpose rep rap the spindle assembly can be totally removed so it wont interfere
with the 3d print, and milling should not get upset by the presence of the extruder.

[Ichan]

How would I go about doing the sums to see what depth of acrylic I should use?

Its a NEMA 17 47nm torque from ebay,  no brand

I do not know any formula to calculate the required thickness of the acrylic 
 
Easy way to understand it's characteristic:
- find a small piece of 6mm thick acrylic
- drill it with 3mm drill bit
- make that 3mm hole larger with 8mm drill bit
- big chance the acrylic will be cracked

Are you sure your Nema 17 stepper motor has 47 NM torque? Or maybe you meant Oz-In?

3D printer can use Nema17 motor because load (extruder) is light weight and it use small timing belt for transmission.

I will suggest to use Nema23 with 200-300 oz-in even for the pcb milling machine, remember with stepper motor better to have extra torque - or missing steps will become a problem.

-ichan

[i_eddy_arv]

I'll give that a shot,
I think I understand what your getting at.   

This will upset my plans, as the rails will have to be a tight fit, meaning they could fracture the acrylic 

I.

[baoshi]

I build one out of the popular MF70 micro mill. http://www.ba0sh1.com/mf70-cnc
Like Ichan said, Nema23 is the least you should go with. Nema17 is too weak for cutting. Use Bi-polar driver for maximum torque at low speed and do not set too much steppers.

[i_eddy_arv]

Thanks baoshi and Ichan

I will upgrade to NEMA23s' I'm also testing some Acrylic samples, and I'll post the results when I get a chance

I.

Posted on: September 17, 2014, 11:00:51 11:00 - Automerged

I do not know any formula to calculate the required thickness of the acrylic 
 
Easy way to understand it's characteristic:
- find a small piece of 6mm thick acrylic
- drill it with 3mm drill bit
- make that 3mm hole larger with 8mm drill bit
- big chance the acrylic will be cracked

Did the test and the Acrylic did not crack at all,
maybe it was the slow drill speed or that the material is quite dense.

BUT Plywood is so much cheaper....

[Gallymimu]

It's hard to have a mill worth a darn without some heavier equipment.

I got a harbor freight conversion special and it's just barely suitable to do light metal work with pretty good accuracy.

http://www.harborfreight.com/two-speed-variable-bench-mill-drill-machine-44991.html
http://www.cncfusion.com/smallmill1.html

plus motors and drivers and pulse generator to interface with Mach 3. 

The tooling (vice, clamps, mounts, measurement tools, bits) cost as much as the whole mill setup.)

[CocaCola]

It's hard to have a mill worth a darn without some heavier equipment.

This is true for metal work, but there are plenty of plastic, wood and light metal CNC mills and routers that do just fine for basic wood working, foam, plastic or even the occasional alloy part...  Not everyone or every part needs precision high tolerances so the flex and slop is small DIY rigs is acceptable many times...

$ for $ if you want to do metal work get an old full sized vertical Bridgeport (or other brand) mill, they can be had used for as low as $500 if you are lucky and don't mind one that is ugly and needs a little TLC...  Or you can get some pretty nice ones for $1000-$1500 if you are patient...  Do your own CNC conversion and you have a rock solid CNC machine for A LOT less then new...

[i_eddy_arv]

3D printer can use Nema17 motor because load (extruder) is light weight and it use small timing belt for transmission.

what about a direct drive ( NEMA Motor drives the shaft directly),  this will give greater torque per step ( as opposed to the timing belt option )
all be it slower, but each pulse will have a small influence on the movement of the axis so a skipped step here and there
will be of very little consequence?

with the timing belt option the distance per pulse is multiplied by the radius or even circumference of the drive wheel, with the direct drive the distance moved per pulse will be the radius or circumference of the axle divided by the thread pitch ( or something close to that )

[snowman]

Hi folks,

I love CNC systems. Here how I made my cnc machine. I started with polywood and dremel rotary tool. I used ordinary long screws and nuts. I used long metal bars in dow matrix printers. And some stepper motors from old printer. I used cheap controller TB6560. you u can find in ebay. and then  I used msch3 to control the board. It was incrediblely precise. mach3 has its own backlash compensation so if you calibrate it right, you'll get to exact same point even after millions of moves.

Then I worked on aluminum on my wooden machine. I made new motor mounts and everything. then by using junk, I had metal cnc machine. I say dont think about too much on stepper powers. if you use leadscrews you'll get more than you need. I never used belt system. I always mouunted motors to leadscrews directly. Now I have a sherline mill. spindle is almost 2kg and z axis works like crazy.

I strongly suggest you to watch videos on youtube. even there's one cnc made from water pipes. You'll be suprised what can be done using broken dot matrix printer parts. After all this years I can say that its not a rocket science. Once you start building it you'll see which parts need to be changed or re-make. even my first wooden model can perfectly drill pcbs.

good luck.

[kreutz]

A few years ago I modified the Sieg X1 micro-mill to CNC. Here is one of the videos: http://www.youtube.com/watch?v=3MdLPhbaVYA
It took a couple of months of free time and several weekends. At the end I had spent about four times the cost of the machine itself in parts and tools, the machine itself was a gift from a friend. Nowadays it is being sitting, unused, in my shop for more than 6 months...

[jellybean442]

Make sure to avoid standard (UNC) threading for the lead screws (if you haven't already). The thread of choice for linear motion is ACME thread, which you can find at most industrial supply houses in a large variety of sizes.

The automation I work with every day utilizes redundant stops, typically a combination of software encoder tracking and/or microswitches, prox switches, photoeyes, etc. Most of the machines also have padded hard stops (anything from pneumatic shocks to rubber pads) to help avoid damage in case of other failures. Better to have a progressive stop versus a sudden, uncontrolled stop.

[Vineyards]

I bought a mini-cnc micro-engraver system thinking I could make prototypes but I could not get it to work.
I am not very good at mechanical things. I messed with g-codes and the stuff but somehow I couldn't get the perspective.

[pickit2]

I bought a mini-cnc micro-engraver system thinking I could make prototypes but I could not get it to work.
I am not very good at mechanical things. I messed with g-codes and the stuff but somehow I couldn't get the perspective.

if your going to try again, work on one axis only, till you get it right.
use a pen and paper before a cutting tool, try marking out, then cutting out front panels is easy to learn.

[i_eddy_arv]

The thread of choice for linear motion is ACME thread,

The automation I work with every day utilizes redundant stops,....
 Better to have a progressive stop versus a sudden, uncontrolled stop.

Totally Agree with all of the above,
I have sourced ACME Threads, and looking at a soft stop mechanism, with stop switches placed well within the field of travel of the axis,  abrupt stops are always bad news especially for rigid assemblies

The chassis has been totally redesigned along with the electronics ( My specialty ) So the project build has been set back by a few months, but for a better mill hopefully.