want to get feet wet... looking for recommendations

[mistermikev]

so I've been thinking a lot about getting my feet wet w cnc.  I'd like to learn to do 3d design, but w/o a way to realize the design it seems a bit futile.  Not looking to take over the world, but would like to be able to dip in my toes and accomplish some small things.

goals:

cut inlay from blanks, cut corresponding inlay channel

make knobs

make pickup covers

budget: I guess I'd like to know what a realistic budget for a small cnc setup is... but is $300 enough to get a small machine?  I could probably spend more, but not looking to invest a lot at this point.

I know I've seen a few threads from @Prostheta here, but have some specific questions and would like recommendations.

1) can you recommend a cnc machine?  don't really want to build as not mech inclined... but am inclined on the computer side.

2) free cad/cam program?

3) thread recommendations?

[MiKro]

Mike to get started plan on about $3000 min.  I'm sure there may be some cheaper but I would not trust them to be accurate enough to do inlays. The only free CAD/Cam is Fusion 360. it is online based and the learning curve is very steep. It will do 3D as well as 2D but again very steep learning curve. Plan on about a year, maybe 6 months if you are working at it 5 to 6 hours a day.

mk

[mistermikev]

28 minutes ago, MiKro said:

Mike to get started plan on about $3000 min.  I'm sure there may be some cheaper but I would not trust them to be accurate enough to do inlays. The only free CAD/Cam is Fusion 360. it is online based and the learning curve is very steep. It will do 3D as well as 2D but again very steep learning curve. Plan on about a year, maybe 6 months if you are working at it 5 to 6 hours a day.

mk

well altho discouraging... I appreciate the advice.  If I may ask a few follow up q's?  so... I understand one can get a larger x carve that is big enough to do a body in the $1500 range... so these are no good?

afa learning curve... I'm sure it would a challenge, and take exactly the time you mentioned... but if I could do it a little at a time I think it would be fun and worth it.  It has taken me years to learn photoshop/gimp, even longer to learn SQL... but with both I was able to do some basic handy things pretty quick... I expect fusion might be similar?  also, are any of the 'hobbyist' licenses worth looking into?

[MiKro]

An xcarve is crap in my opinion. Now that being said that does not mean you cannot do what you want with it. It will require a lot more sanding on your part and don't expect it to be super accurate. As far as hobby licenses, no, many do have educational license though. Fusion 360 is free from Autocad. Vectric does have Cut2d, it is a very basic 2d CAD/CAM and around $300. Some use Easel which is a 2d cad it is cranky at best.

[mistermikev]

29 minutes ago, MiKro said:

An xcarve is crap in my opinion. Now that being said that does not mean you cannot do what you want with it. It will require a lot more sanding on your part and don't expect it to be super accurate. As far as hobby licenses, no, many do have educational license though. Fusion 360 is free from Autocad. Vectric does have Cut2d, it is a very basic 2d CAD/CAM and around $300. Some use Easel which is a 2d cad it is cranky at best.

Thank you very much for your input.  I def appreciate it. 

ok, so xcarve is no good.  is it better to buy a kit an assemble yourself, otherwise can you do a light recommendation afa "just something to not expect much from but a good introduction tool"?

[curtisa]

I'd actually suggest that for the small items you've nominated it is entirely possible to do it on a budget CNC. I got started on a prefab Chinese kit from eBay, a CNC3020. The model number just refers to the rough dimensions of travel; in this case 30cm x 20cm. From memory it was under $700AU at the time, but for more money I could've gone with bigger models - CNC3040, CNC6040, CNC9060 etc. They all tend to be pretty similar in terms of quality, construction and implementation. I used the 3020 unit for several years and had it do fret slotting and inlays just fine. To be fair though, that's pretty light duty work and probably all that you could expect for such a small, cheap machine. It certainly couldn't cope with milling metal or deep cuts in timber.

The drawback to going with a cheap all-in-one unit is that sooner or later you will run into its limitations - the cutting area will eventually be too small, there'll be too much runout in the spindle, the frame and gantry will flex too much. steppers will be too weak, spindle can't take the bit you're wanting to use, not enough vertical travel etc. And if you want to venture deeper down the CNC rabbit hole, you'll invariably outgrow the basic unit and have to shell out extra for either a more mature kit or build your own. The $3K figure that @MiKro suggests is probably about what you'd plonk down for that 'tier 2' unit.

The software will also require a significant investment on your part in terms of training and time. You're essentially looking at working in three or four different headspaces concurrently - Computer Aided Drafting, 3D Modelling, Computer Aided Manufacturing and Motion Control of the CNC hardware itself. Some software packages attempt to integrate the lot. Others will only perform one or two of those functions.

There are other options regarding software for the CNC. Freecad has integrated CAD, 3D modeling and CAM elements and is constantly being expanded and developed as time goes by. Being open source it is free, and for the most part it does the job. Make of that what you will . At the low-cost end of the market QCAD has decent built-in CAM capabilities and has a much better implementation of CAD than Freecad currently has, but is strictly 2.5D at this stage. Although to be honest, on a small CNC machine you probably woudn't attempt any true 3D milling anyway.

For actually driving the CNC you only have a couple of options - Mach 3/4 ($$) or LinuxCNC (free). Having gone down the LinuxCNC path I actually wouldn't recommend it to absolute beginners as it can be an absolute pig to get working and the support for it is somewhat lacking. If you're used to working with Linux computer environments then you may get along with it OK, but I've personally found it's implementation, usage and technicalities pretty, well...Linux-y.

If you have moderate DIY electronics skills, what does seem to be a good entry point into the CNC motion control system is the use of an Arduino running the GRBL firmware. I don't think there are any decent fully-fledged CNC kits that use it yet, so you'd have to retrofit it intoyour own hardware, but for basic XYZ CNC open-source machines it is an attractive alternative and has a lot of things going for it if you don't need the complexity of LinuxCNC or Mach.

[mistermikev]

4 hours ago, curtisa said:

I'd actually suggest that for the small items you've nominated it is entirely possible to do it on a budget CNC. I got started on a prefab Chinese kit from eBay, a CNC3020. The model number just refers to the rough dimensions of travel; in this case 30cm x 20cm. From memory it was under $700AU at the time, but for more money I could've gone with bigger models - CNC3040, CNC6040, CNC9060 etc. They all tend to be pretty similar in terms of quality, construction and implementation. I used the 3020 unit for several years and had it do fret slotting and inlays just fine. To be fair though, that's pretty light duty work and probably all that you could expect for such a small, cheap machine. It certainly couldn't cope with milling metal or deep cuts in timber.

The drawback to going with a cheap all-in-one unit is that sooner or later you will run into its limitations - the cutting area will eventually be too small, there'll be too much runout in the spindle, the frame and gantry will flex too much. steppers will be too weak, spindle can't take the bit you're wanting to use, not enough vertical travel etc. And if you want to venture deeper down the CNC rabbit hole, you'll invariably outgrow the basic unit and have to shell out extra for either a more mature kit or build your own. The $3K figure that @MiKro suggests is probably about what you'd plonk down for that 'tier 2' unit.

The software will also require a significant investment on your part in terms of training and time. You're essentially looking at working in three or four different headspaces concurrently - Computer Aided Drafting, 3D Modelling, Computer Aided Manufacturing and Motion Control of the CNC hardware itself. Some software packages attempt to integrate the lot. Others will only perform one or two of those functions.

There are other options regarding software for the CNC. Freecad has integrated CAD, 3D modeling and CAM elements and is constantly being expanded and developed as time goes by. Being open source it is free, and for the most part it does the job. Make of that what you will . At the low-cost end of the market QCAD has decent built-in CAM capabilities and has a much better implementation of CAD than Freecad currently has, but is strictly 2.5D at this stage. Although to be honest, on a small CNC machine you probably woudn't attempt any true 3D milling anyway.

For actually driving the CNC you only have a couple of options - Mach 3/4 ($$) or LinuxCNC (free). Having gone down the LinuxCNC path I actually wouldn't recommend it to absolute beginners as it can be an absolute pig to get working and the support for it is somewhat lacking. If you're used to working with Linux computer environments then you may get along with it OK, but I've personally found it's implementation, usage and technicalities pretty, well...Linux-y.

If you have moderate DIY electronics skills, what does seem to be a good entry point into the CNC motion control system is the use of an Arduino running the GRBL firmware. I don't think there are any decent fully-fledged CNC kits that use it yet, so you'd have to retrofit it intoyour own hardware, but for basic XYZ CNC open-source machines it is an attractive alternative and has a lot of things going for it if you don't need the complexity of LinuxCNC or Mach.

some great info there - thanks for that.  how did you do fretslots on something that is that small?  I guess if you lied diagonal it was just enough space?  either way, that'd be beyond my expectations... looks like a good option for getting in the door.  $500 and I could build a linux box to control it.  would have to dust off my linux shoes but that'd probably be good for me.  man pages don't scare me.

I see the machines you mention on evilbay.  something to dig into there. 

A wealth of info here and a lot to digest, will have to go do some research on a few of these techs.  Thank you for the responses!

 

[MiKro]

Thanks Curtisa, I don't know enough about the Chinese CNC machines. I was looking into them for Mike so I could maybe point him in a direction.

MK

[curtisa]

16 hours ago, mistermikev said:

how did you do fretslots on something that is that small?  I guess if you lied diagonal it was just enough space?

By a process known as tiling. The fretboard gets attached to a substrate that has some locating pins milled into it that allows it to be moved forwards and backwards by a known amount. The fretboard gets milled in two halves by first attaching the substrate into key position 1 and milling the first section of the fretboard, and then moving it forwards to key position 2 for the second half of the job.

I wrote a series of articles on the process for Projectguitar some time back which can be found in the Articles Section of the website. There are also other articles where I documented my experience with one of those cheap Chinese mini CNCs.

The other way you can employ a mini CNC is to use it to create router templates rather than directly milling something. Cutting a pickup cavity into a body is impractical on the mini CNC as it either won't fit on the bed, or the spindle won't cope with the deep cuts required. Milling a pickup cavity router template out of 1/4" MDF however is perfectly achievable. That opens its usefuness up by a significant margin, as you can now make cavity covers that precisely match a template, create a series of interlocking custom templates for a Floyd Rose trem cavity, drilling templates for multiscale bridge assembly, or mill a scratchplate out of blank pickguard material etc

[sprack]

You can do it in the vicinity of $1.5k. I'm just in just a little over $2000 USD in the hole with a Shapeoko XXL + accessories.

CAD/CAM in Fusion360. Remote controlling it via a raspi 3+ with CNCJS, which I found much easier than LinuxCNC.

 

[mistermikev]

5 hours ago, curtisa said:

By a process known as tiling. The fretboard gets attached to a substrate that has some locating pins milled into it that allows it to be moved forwards and backwards by a known amount. The fretboard gets milled in two halves by first attaching the substrate into key position 1 and milling the first section of the fretboard, and then moving it forwards to key position 2 for the second half of the job.

I wrote a series of articles on the process for Projectguitar some time back which can be found in the Articles Section of the website. There are also other articles where I documented my experience with one of those cheap Chinese mini CNCs.

The other way you can employ a mini CNC is to use it to create router templates rather than directly milling something. Cutting a pickup cavity into a body is impractical on the mini CNC as it either won't fit on the bed, or the spindle won't cope with the deep cuts required. Milling a pickup cavity router template out of 1/4" MDF however is perfectly achievable. That opens its usefuness up by a significant margin, as you can now make cavity covers that precisely match a template, create a series of interlocking custom templates for a Floyd Rose trem cavity, drilling templates for multiscale bridge assembly, or mill a scratchplate out of blank pickguard material etc

ahhh, I could see how that could work.  very cool.  will have to go read through those threads. 

I could see a lot of things that could be done there, control cavity covers/templates, smaller pickguards... also I do a lot of prototype pcbs and would love to automate the drilling of holes.  I imagine you could even forgo etching and route the pcb traces... but that actually might be too much precision/work.

thank you so much for the wealth of info.  you are a real resource here.

 

4 hours ago, sprack said:

You can do it in the vicinity of $1.5k. I'm just in just a little over $2000 USD in the hole with a Shapeoko XXL + accessories.

CAD/CAM in Fusion360. Remote controlling it via a raspi 3+ with CNCJS, which I found much easier than LinuxCNC.

 

thank s for the info.  your workshop is impressive clean/organized!

[curtisa]

5 minutes ago, mistermikev said:

I imagine you could even forgo etching and route the pcb traces... but that actually might be too much precision/work

Nope. That's also a perfect task for CNC. Protoype PCBs are nothing more than fancy engraving jobs, and there are several free solutions floating around for taking gerbers and generating milling code.

[mistermikev]

2 minutes ago, curtisa said:

Nope. That's also a perfect task for CNC. Protoype PCBs are nothing more than fancy engraving jobs, and there are several free solutions floating around for taking gerbers and generating milling code.

good to know.  that would be huge for me.  no more dangerous chems and tedious drilling of holes!

[mistermikev]

just wanted to thank everyone again as I'm in here re-reading this thread... lots of helpful info - thanks again for that.

[MiKro]

@mistermikev

I was going through this thread looking at everything again.

I was asked by someone on FB if I could design an affordable CNC for milling bodies and possibly basic necks.

With what I am working on I think I can make a machine complete that will have a working cut area of approx 32"x 20" x 6" for under $1000. It would include a VFD with spindle control, ball screws and linear guides.  No software though. If this pans out, (hopefully?) I hope to have at minimum accuracy in the +/- 0.005" range or better. Not great for neck fretting ( no fretboard slotting unless the accuracy is below +/- 0.002") but would do bodies and basic necks with no problem as wood will move this much due to humidity changes anyway. Will keep everyone posted. This my take a while. LOL!!

Man I hate to put 1k in to an unknown project. LOL!!

MK

[curtisa]

Sounds like an interesting project, Mike. I take it this will be MDF construction (or similar) rather than aluminium/steel?

8 hours ago, MiKro said:

I hope to have at minimum accuracy in the +/- 0.005" range or better. Not great for neck fretting ( no fretboard slotting unless the accuracy is below +/- 0.002")

0.005" is a little over a tenth of a mm. I would think that'd be better accuracy than you could achieve with a hand slotting mitre box and template. Once you factor in dressing/re-crowning the frets, playing style, seasonal movement of the materials etc, it's unlikely that even a base accuracy tolerance of twice that value would cause much concern in practice.

[MiKro]

9 hours ago, curtisa said:

Sounds like an interesting project, Mike. I take it this will be MDF construction (or similar) rather than aluminium/steel?

0.005" is a little over a tenth of a mm. I would think that'd be better accuracy than you could achieve with a hand slotting mitre box and template. Once you factor in dressing/re-crowning the frets, playing style, seasonal movement of the materials etc, it's unlikely that even a base accuracy tolerance of twice that value would cause much concern in practice.

I agree but with reservations on the accuracy?

As far as construction, I am playing with the idea of a High Grade Baltic birch ply with Hard maple  or aluminum , maybe steel bar reinforcement, using a torsion box type construction for the bed. This coupled with Steel Linear Round rails and bearings vs. the v groove type should enhance the rigidity of it overall.

Will see what happens

MK

 

bearing type I am looking at.

 

 

Beam style with angle inserts

[mistermikev]

18 hours ago, MiKro said:

@mistermikev

I was going through this thread looking at everything again.

I was asked by someone on FB if I could design an affordable CNC for milling bodies and possibly basic necks.

With what I am working on I think I can make a machine complete that will have a working cut area of approx 32"x 20" x 6" for under $1000. It would include a VFD with spindle control, ball screws and linear guides.  No software though. If this pans out, (hopefully?) I hope to have at minimum accuracy in the +/- 0.005" range or better. Not great for neck fretting ( no fretboard slotting unless the accuracy is below +/- 0.002") but would do bodies and basic necks with no problem as wood will move this much due to humidity changes anyway. Will keep everyone posted. This my take a while. LOL!!

Man I hate to put 1k in to an unknown project. LOL!!

MK

right on.  sounds like some interesting stuff altho half of it is over my head. 

for me, and at this point... I'm leaning towards something somewhat 'disposable'.  My budget for entry is $3-500.  Something I can try it out and not be too concerned with tossing it down the line.  I figure, worst case scenario... I would have a dedicated pcb fabrication tool.  I just want to get my feet wet and see how much work it is to do some basic stuff like cutting a headstock inlay or doing control covers along with being able to drill pcbs.  I dunno how realistic it is because I can't see myself building the same thing more than 1 or 2 times.  likely the work to accomplish that on cnc would make it impractical.  That said, if I find it isn't that much work to make a body template (for instance), my plan is to invest more in it at that point... and then I would potentially be willing to put $2k-4k in.     

Really still in the 'exploratory' stages at this point... so all of this could change.  I keep reading about it here and there and thinking I should just buy something cheap to get the ball rolling... but I know what I really should do is do some 3d design work on one of the free tools to get going.

[MiKro]

13 minutes ago, mistermikev said:

right on.  sounds like some interesting stuff altho half of it is over my head. 

for me, and at this point... I'm leaning towards something somewhat 'disposable'.  My budget for entry is $3-500.  Something I can try it out and not be too concerned with tossing it down the line.  I figure, worst case scenario... I would have a dedicated pcb fabrication tool.  I just want to get my feet wet and see how much work it is to do some basic stuff like cutting a headstock inlay or doing control covers along with being able to drill pcbs.  I dunno how realistic it is because I can't see myself building the same thing more than 1 or 2 times.  likely the work to accomplish that on cnc would make it impractical.  That said, if I find it isn't that much work to make a body template (for instance), my plan is to invest more in it at that point... and then I would potentially be willing to put $2k-4k in.     

Really still in the 'exploratory' stages at this point... so all of this could change.  I keep reading about it here and there and thinking I should just buy something cheap to get the ball rolling... but I know what I really should do is do some 3d design work on one of the free tools to get going.

Mike, this may lead to also a smaller type for what you are trying to do. I also have a friend that wants one for pcb only work. Mainly about a 10" x 10" x 3" work area.

[MiKro]

I think maybe some steel bed frame angle would be possible. These can be had at many thrift stores for almost nothing. One bed frame cut up could add the extra rigidity to this at a very reasonable cost.

If I use a router based system vs a spindle it should cut the cost by another $150 most likely as well.

The goal for me would be to make a solid first cnc that someone could build for guitar and other work for less than 1k not including design software.

There are some free CAD/CAM softwares. like Fusion 360 as one. While Linux cnc controller software is free it is as suggested above a little clunky for beginners. Mach3 though is less than $200. Will see where this goes. Now to design one in CAD first.

 

MK

something like this for the steel angle material. Cut it up to needed lengths

 

[mistermikev]

33 minutes ago, MiKro said:

Mike, this may lead to also a smaller type for what you are trying to do. I also have a friend that wants one for pcb only work. Mainly about a 10" x 10" x 3" work area.

Right on.  bed frame - good thinking there.

[MiKro]

1 minute ago, mistermikev said:

Right on.  bed frame - good thinking there.

Something like this for the beams?

 

[MiKro]

Could also drill and tap the steel to mount the Linear rails.

Obviously the bed frame material would be inset , epoxied and screwed to the baltic birch.

MK

[mistermikev]

27 minutes ago, MiKro said:

Could also drill and tap the steel to mount the Linear rails.

Obviously the bed frame material would be inset , epoxied and screwed to the baltic birch.

MK

right o.  the other day I watched a build thread of one of the cnc kits... the workbee.  Was surprising how straight fwd that actually was.  still - have to do all the work and the kit is over a grand.

[MiKro]

31 minutes ago, mistermikev said:

right o.  the other day I watched a build thread of one of the cnc kits... the workbee.  Was surprising how straight fwd that actually was.  still - have to do all the work and the kit is over a grand.

I would bet that my design will be more robust than a workbee. LOL!!!

While the workbee only  uses V-type linear rails, as well as lead screws and some other (what I consider substandard parts). Yes it is an aluminum frame but I  think a good solid baltic birch with inset steel will be as strong. I also will be using much better Ball screws, and roller bearing linear rails.

It's all in my head though at the moment, so in reality I need to get the design down and build one. LOL !! Not that I need one for me but it is a challenge.