Small CNC Router Build

[worn thumbs]

Congratulations on reaching a landmark stage,its rather special to see the machine moving isn't it?Calibrating and setting home/limit switches is perhaps a bit less epic,but no less satisfying-its also something I have yet to do with my machine although with ballscrews the calibration was easy.

Do you have any software for generating toolpaths yet?I have been using Freecad's path module and it does a very good job for most basic shapes.The attached message was created using F-engrave and since it too is free,you might find it worth a play.It works with bitmap images or .dxf files.

 

[AES]

That looks really good worn thumbs - especially so when one considers that it comes from a machine that you've built and programmed yourself. Must be a very satisfying feeling.

As above, "I takes me 'at orf" to all you clever so n so s.

 

[CHJ]

Just as a matter of personal interest, no matter how good it is mechanically, there must be some backlash, especially when the cutter changes axis in a 3D cut. How is that allowed for? I presume some software "tweak"? (Or have I completely misunderstood - quite likely)?

Not a lot of backlash in the drive system supprisingly stiff, I guess the software takes care of the rest.

Be some time before I get anywhere near this:-
https://www.youtube.com/watch?v=-Y0OHRD ... fUzR-vqB6o

 

[CHJ]

Do you have any software for generating toolpaths yet?I have been using Freecad's path module and it does a very good job for most basic shapes.The attached message was created using F-engrave and since it too is free,you might find it worth a play.It works with bitmap images or .dxf files.

I have some at hand, thanks for the tips, I am holding off delving into which software to use until I can get to fitting a Router Motor and run some exsiting Gcode.
I envioson taking the simplest route possible for mainly surface designs, possibly a little Z axis contouring.

Thanks for the tips regarding Freecad and Engrave, have played with software associated with 3D printer that was used to print the parts for me so have an inkling of what's needed to generate to G Code.

 

[CHJ]

...Just as a matter of personal interest, no matter how good it is mechanically, there must be some backlash, especially when the cutter changes axis in a 3D cut. How is that allowed for? I presume some software "tweak"

Yer 't'is, one of many parameters that can be tweaked

Who knows when, if even, I can manage to get my head round them all.

 

[AES]

Ah, OK Chas. At least I understood the bit in the table about back lash! I venture to think that it will be an interesting exercise to measure the back lash so you know by what amount to tweak. Though thinking about that a little more, I guess the easiest way will be to just fiddle with a test cuts piece until you get the value right? I wonder if that value will change though, dependant on the cutter and the density of the wood you're cutting? Interesting - but I must be careful not to strain my few brain cells - have you seen my (quite ridiculous) post about 2018 in Off Topic???

 

[CHJ]

John, (wt)
What spindle motor are you using, and specifically the wattage for your sample above.

I only expect to cut wood patterns and simple mouldings with it but really have no correlation as to wattage required for something not much greater than engraving compared with say hand held routers.

Envisioned something along these lines may be adequate for my needs

 

[worn thumbs]

Since my machine was intended as a learning exercise in CNC building and I had no plans to produce chronometer parts or space shuttles,I kept it basic.I happened to have a Rotozip that I bought when B&Q had them on clearance at less than £40 and as far as I recall it is a 550W motor.I wouldn't want any less as it is the limiting factor with this machine.Several of the home built machines on youtube feature the Bosch palm router and I can see these being reasonable.In view of the likely shortage of Rotozip spares I made a move to future proof myself by buying an Aldi palm router a few weeks ago-500W and even a couple of cutters for £25 and I have no doubt it will work if needed.

The dedicated spindles in your link look promising and If (when!) I move on to a bigger and better machine they are the sort of thing I would want.For strictly small scale work you wouldn't need it but I would want to have at least an ER16 collet so that 1/2" shank cutters are feasible.The extra cost of the hardware would soon be overtaken by all the hours a small motor tickling the surface would be consuming electricity.The best analogy is to think of the effort and number of cuts it would take something like a small 1/4" router to do a job that a 1/2" router could do in one cut.I would guess that for solely engraving you could get away with 200W.

I intend to spend some time over the next few days learning more about home and limit switches as I currently operate without them.I bought both the contact type and the inductance type and while the commercial machines I have used had the inductance type,they do seem to require quite a range of movement to establish a home position and this eats into the work envelope.I have done a little reading and early signs are that homing and limits with LinuxCNC are not the easiest of topics to master.

For now I am delighted to be able to produce accurate parts (by woodworkers definition) and to avoid some of the tedious jobs of template making that I had ceased to enjoy.Its also fun to V-carve house numbers and the like.

 

[CHJ]

Another major step forward, checked movement calibration, Z axis was a factor of 4 too high so had Marlin software modified for me, installed Arduino on PC and received instruction on how to do it myself in future, and uploaded modified firmware to Board, Axis movement now calibrated. The wonders of Skype.

 

[worn thumbs]

Another major step forward, checked movement calibration, Z axis was a factor of 4 too high so had Marlin software modified for me, installed Arduino on PC and received instruction on how to do it myself in future, and uploaded modified firmware to Board, Axis movement now calibrated. The wonders of Skype.

That would be a very useful stage completed.Does the change to the Z axis now mean that the resolution has improved by a factor of 4 too?My steppers have 200 steps per rotation and with a 5mm pitch on the ballscrews,that gives a resolution of 1 step=1/40th of a mm which ought to be more than enough for most woodworking purposes.It is possible to adjust the dip switches on the control modules to permit micro-stepping to something like one sixteenth of that-the penalty being a loss of torque.

 

[custard]

This is such a great project, thanks for posting!

It also looks like you've found a real sweet spot. Big enough to be genuinely useful, yet small enough to be manageable.

=D>

 

[CHJ]

…..That would be a very useful stage completed.Does the change to the Z axis now mean that the resolution has improved by a factor of 4 too?.......

This was just distance travelled from a given Manual input of the Java CNC-GCode-Controller program.
Set tabs of 0.1 1.0 5.0 10.0 mm (I.E. 10mm was giving me 40mm travel)

Have not worked out what the actual resolution is on the steppers, I am using 16 tooth T2 pulleys on Nema 17 (200 step) steppers.
X & Y are belt transmission, Z is 8 x 2 mm screw with a small gear ratio between.

 

[CHJ]

This is such a great project, thanks for posting!

It also looks like you've found a real sweet spot. Big enough to be genuinely useful, yet small enough to be manageable.

=D>

Thanks Custard, Hopefully I'll see it through to actually cutting wood sometime in the next year, realistically it's just a slow burner project in an effort to keep the old grey cells working.

 

[worn thumbs]

A 2mm thread pitch with 200 steps/rotation would give fantastic resolution and the gearing would have to be fairly severe to degrade it.

 

[CHJ]

We have following of Gcode, of sorts having played today with setting start code and end codes to put things where they need to be to clear work area etc.

Main thing now is to generate some better GCode samples to play with and sort out Z movement positioning.

Who knows the new year may see enough progress to contemplate a Spindle motor acquisition.

 

[Steve Jones]

Hi Chas

Just to put my two 'penneth in re the spindle

When I built my machine a couple of years ago which admittedly is quite a bit bigger than yours I installed a Bosch POF500 router, this didn't last long so it was replaced with a Trend T3 which I replaced the main bearings in twice !
The draw back for me with these type of spindles is noise, so I changed to a 1.5kw air cooled spindle from eBay, the difference is chalk and cheese, although there's still noise it's not the high pitched squeal put out by the routers and I can now hold a conversation while I'm cutting something. I have seen the spindles you linked to in the flesh so to speak as a mate bought one for testing purposes and they're not bad at all.

cheers

 

[CHJ]

.. I have seen the spindles you linked to in the flesh so to speak as a mate bought one for testing purposes and they're not bad at all.

Thanks for the input Steve, They are the pattern that the existing printed clamps are sized to hold and I doubt the assembly is capable of sensibly holding or driving anything physically 'bigger', they should be more than good enough for what I envision doing.

 

[CHJ]

Looks like we may be getting somewhere.

 

[AES]

Yes Chas, "we" are definitely getting somewhere! I've been following this thread with great interest, not because I'll ever try such a thing myself (I really don't understand the PC stuff, and frankly, am too stupid & lazy to try) but A) because I can at least understand and admire the electro-mechanical stuff, and B) because just as custard has said, you're going to end up with something eminently practical - and all "with your own hands"! I won't use the word "awe" (much over-used IMO) but I will definitely admire!

Happy New Year and thanks for posting.

 

[CHJ]

Still a lot to get the brain cells reactivated about yet Andy.

As well as researching the process specific jargon and finding out how to set or adjust settings in the controlling software, I need to sort out a design/conversion software that will do the simple stuff I need, don't want to go through the same sort of learning curves I did 40-50 yrs. ago. I'll see what I can learn from what I've got first.

Example, not yet having mastered precis levelling, the bottom row of text failure at the end was due to the difference of the two layers of packing paper on the left.
Wood carving will not care about such precision so should be able to ignore that aspect.

May get away with sorting out file editing and saving software that I can use to modify existing designs to suit my needs.