want to get feet wet... looking for recommendations

[mistermikev]

1 hour ago, MiKro said:

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.

right on... wasn't necc making comparison, just was surprised that it wasn't as complicated as I thought to assemble.  I think for me... whatever I buy... not interested in doing the assembly.  At least not at this point.

[MiKro]

So I ordered the linear rails and ball screws today. While I know what the measurements are, I like to have things in hand for me to make a design of this nature.

Sorry, I don't trust printed specs much anymore. LOL!!!

So far this is $256.78.

mk

update: so the sizes are this.

1000 mm rails = 39.37 inches

600 mm rails = 23.62 inches

300 mm = 11.8 inches

This should lead to approx 36" x 20" x 5" cut area. More than plenty for most guitars and necks.

[ScottR]

11 hours ago, MiKro said:

Sorry, I don't trust printed specs much anymore. LOL!!!

Amen brother!

SR

[MiKro]

So looking at different electronics. It appears that most of what I need will come in around $350.

This will put the major parts at about $600 total for the cnc minus a few things. Still need a sheet of 3/4" baltic birch, A low end PC with Xp or win7, screws and some other odds and end stuff.

So still on track for  $1k or under.

If all goes well I expect rapids in the 200-300IPM range with this and the accuracy under 0.005"+/-

mk

If nothing else when this is done I can attach a laser head to it instead of router and have a larger than normal Laser engraver. LOL!! I can get a Diode head cheap. The stepper motor size will be overkill for a laser though. LOL

[curtisa]

What are your plans for the electronics?

[MiKro]

19 hours ago, curtisa said:

What are your plans for the electronics?

@curtisa

Parallel port type Break out board with PWM control, Keling digital drivers, 420/oz nema23 steppers dual shaft. 40volt PSU for drivers and a 12volt/5volt PSU for the PWM/BOB .  I'm trying to keep it low cost. So no Smoothstepper or higher end PC.  This will allow using mach3 or Linux CNC for controller. Must adhere to good shielded cables so as not to introduce any RFI problems. By using the dual shaft Steppers, I can also add/make dampers that will offset any midband resonance. This combined with the 40 volt PSU should easily get 250IPM or greater with out problems.

While I will be using a parallel port, I can only run at a max frequency of 35khz maybe a hair more. Unlike a higher end PC and ESS Smoothstepper using WIn10 /ethernet where I could go to 100khz with no problem and get more speed without losing too much torque.

Also understand the costs I am using is actual cost, as these figures include any shipping and taxes as well.

While looking at spindle options using a router vs VFD controlled the $$ are marginal. A PC 690R motor for instance is around $200 new. while a 1.5KW or possibly a 2.2KW spindle and VFD can be had for around $230-$275.  Advantage of the VFD type spindles is they can use ER11 for the 1.5KW or ER20 for the 2.2KW collets, which gives a lot more accuracy and choices of end mill sizes. These are the only things that may put this over the $1k mark I am trying to achieve?

 

mk

[curtisa]

You wouldn't consider a more basic electronic backbone such as GRBL? Parallel port control from the PC opens up a lot of flexibiity, but there are good reasons to steer clear of it for an entry-level CNC. It's an antiquated connection that is getting more difficult to find on a new PC let alone a laptop. I'm not familiar with Mach 3 but LinuxCNC is notoriously picky with regards to PCs and parallel ports - just because the PC has one fitted still doesn't mean it will be suitable for motion control, and is the sort of thing you only find out about if you try running the software and/or are willing to spend a lot of time making OS and BIOS tweaks. USB-parallel adaptors won't work with LinuxCNC either, so you're really limited to researching for an exact PC hardware setup that is known to work, or crossing your fingers and hoping that whatever box you have on hand is up to the task.

GRBL is platform-independent and runs on an Arduino which can be had for a few bucks each. The GRBL firmware itself is free, as is the development environment that is used to configure it and upload it to the Arduino. The motion software on the host computer (again, there are decent free options) just connects through any USB port and streams the G-code to the Arduino, letting it parse each line by itself.  The IO on the Arduino is 5V tolerant, so it's still compatible with just about any step/direction drive out there. Spindle control is available as a PWM output, plus there are inputs for home/limit switches, e-stop and a probe without requiring a breakout board. For a basic system it's actually got a lot of pluses going for it, particularly if the goal is to make something that is more or less plug-and-play with minimal futzing around.

[MiKro]

@curtisa

I appreciate your thoughts on this very much. While I agree that GRBL would work? My experience with it is from 100's of people using it in the laser world. The fact that USB can be flaky and the problems of motion control in that environment steers me away from it. I also see similar problems with others using USB in CNC so I am one of those on an opposite fence from you in it's use. Thanks for your knowledgeable input and insight on this. It can always be an alternative to my method for some to cut cost. :)

Sincerely,

Mike

 

[MiKro]

update

Two boxes came in today . Linear rails and ball screws. this is the big box. I am also not totally sold on the round type linear bearings though these I ordered are solid support mounted vs being suspended at the ends. I used that type in my Zaxis on my machine and have to make a new one to eliminate the small amount of flex they have due to the suspended at each end configuration, which causes some slight accuracy issues when milling aluminum.

One thing I see is they took care of making sure the mounting holes for the rails are even and the lengths are the same. This will make modelling it much easier for plans. I don't like the couplers as they are weak and will not last  long.

The ball screws and nuts appear to be of good quality and tight. That will assist in the anti-backlash department.

 

 

[curtisa]

2 hours ago, MiKro said:

I appreciate your thoughts on this very much. While I agree that GRBL would work? My experience with it is from 100's of people using it in the laser world. The fact that USB can be flaky and the problems of motion control in that environment steers me away from it. I also see similar problems with others using USB in CNC so I am one of those on an opposite fence from you in it's use. Thanks for your knowledgeable input and insight on this. It can always be an alternative to my method for some to cut cost. :)

No problem Just throwing out another option that keeps things in the 'low cost' bracket.

Don't get me wrong through - My main machine is running LinuxCNC (originally via parallel port, but now through ethernet), but I do have a laser engraver here running GRBL too. While I wouldn't waste time downgrading the bigrig to run on GRBL, there's no doubt that the laser engraver was sooooo much easier to get running from scratch.

Carry on

[mistermikev]

so, if I may...

goal: pcb, headstock template, ornamentals/inlay, pickup/control-cavity/neck-pocket template.

budget: throw away $600 or less

are any of these worth it... and if not is there any specific alternative you could recommend?

3040 - $565

https://www.amazon.com/VEVOR-Engraver-Engraving-Drilling-Cutter（3/dp/B08FQRZ4CJ/ref=sr_1_1_sspa?dchild=1&keywords=cnc+3040&qid=1601464106&sr=8-1-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUExWFRDTFJVQUYwME1UJmVuY3J5cHRlZElkPUEwNTQxOTcwMUYwWEs4S0NMTDRNWCZlbmNyeXB0ZWRBZElkPUEwMTE4NDYyMlowR0E4TkVNVDNOViZ3aWRnZXROYW1lPXNwX2F0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

3018 - $250

https://www.amazon.com/gp/product/B07P6K9BL3/ref=crt_ewc_title_dp_2?ie=UTF8&psc=1&smid=A1CJB5SYI9X4XC

9" x 14" x 2" - $450

https://www.whittlecnc.com/

[curtisa]

The first one is basically the next size up from the unit I got started with, a CNC3020. I see at the bottom of the listing they even show the CNC3020 as still being for available. The controller looks like it's been upgraded and has dropped the parallel port interface that mine came with for USB instead. It's not clear what the control interface actually is, but like the other two units you list it could be a derivative of GRBL/Arduino (plenty of open-source software options) and if so it should just plug-and-play from an interfacing perspective. But otherwise it looks the same as my old machine, just bigger.

The second unit is slightly different in that the gantry is fixed and the table moves underneath it to give the required Y-axis travel, as opposed to the CNC3040 and WhittleCNC where the table is fixed and the gantry moves over it. It makes the overall footprint bigger than the equivalent-sized moving gantry unit, but if its been designed properly it should be more rigid. I doubt that translates to anything meaningful though, as it's very much a light duty machine and you won't be milling metal or taking massive cuts on the 3018. The spindle motor is tiny, so any work you put it to will need to be done on the slow or taking very small nibbles - OK for PCBs and shallow engraving jobs, but expect really long work times for half-inch MDF templates. There's tons of reviews and demos on Youtube for the 3018, so evidently it's got quite a strong following.

I'd probably steer clear of the WhittleCNC. The ability to lower the table to give more potential work clearance is nice, but the whole thing is a very open-to-the-elements design. You're forever going to be cleaning it just to keep it running smoothly and reliably. The spindle motor seems to be some kind of handheld flexishaft engraver? Direct-drive belts and plastic frame mean its really only meant to be used for the bedroom hobbyist.

If it were me I'd go for the CNC3040, if only because it's the one I'm familiar with. Beyond that there are other things about it which float my boat - the spindle is bigger (you'll run it on maximum speed for all your jobs anyway) and will take more abuse, the construction has a lot more mass to it which translates to being able to cope with bigger workloads, the inherent design shields the moving components reasonably well from dust and chips. If nothing else the bigger working area should sway you towards it. You might only foreshadow making up PCBs and pickup cavity routing templates, but I can guarantee there will come a time in the future when you need to cut a custom Strat pickguard and it will not fit no matter how you position it. If you're quite sure you don't need to go as big and are looking to save a few bucks, you can always shoot for the 3020 (which is also a more on-par comparison to the WhittleCNC and 3018 you linked to in terms of physical dimensions anyway).

[mistermikev]

19 minutes ago, curtisa said:

The first one is basically the next size up from the unit I got started with, a CNC3020. I see at the bottom of the listing they even show the CNC3020 as still being for available. The controller looks like it's been upgraded and has dropped the parallel port interface that mine came with for USB instead. It's not clear what the control interface actually is, but like the other two units you list it could be a derivative of GRBL/Arduino (plenty of open-source software options) and if so it should just plug-and-play from an interfacing perspective. But otherwise it looks the same as my old machine, just bigger.

The second unit is slightly different in that the gantry is fixed and the table moves underneath it to give the required Y-axis travel, as opposed to the CNC3040 and WhittleCNC where the table is fixed and the gantry moves over it. It makes the overall footprint bigger than the equivalent-sized moving gantry unit, but if its been designed properly it should be more rigid. I doubt that translates to anything meaningful though, as it's very much a light duty machine and you won't be milling metal or taking massive cuts on the 3018. The spindle motor is tiny, so any work you put it to will need to be done on the slow or taking very small nibbles - OK for PCBs and shallow engraving jobs, but expect really long work times for half-inch MDF templates. There's tons of reviews and demos on Youtube for the 3018, so evidently it's got quite a strong following.

I'd probably steer clear of the WhittleCNC. The ability to lower the table to give more potential work clearance is nice, but the whole thing is a very open-to-the-elements design. You're forever going to be cleaning it just to keep it running smoothly and reliably. The spindle motor seems to be some kind of handheld flexishaft engraver? Direct-drive belts and plastic frame mean its really only meant to be used for the bedroom hobbyist.

If it were me I'd go for the CNC3040, if only because it's the one I'm familiar with. Beyond that there are other things about it which float my boat - the spindle is bigger (you'll run it on maximum speed for all your jobs anyway) and will take more abuse, the construction has a lot more mass to it which translates to being able to cope with bigger workloads, the inherent design shields the moving components reasonably well from dust and chips. If nothing else the bigger working area should sway you towards it. You might only foreshadow making up PCBs and pickup cavity routing templates, but I can guarantee there will come a time in the future when you need to cut a custom Strat pickguard and it will not fit no matter how you position it. If you're quite sure you don't need to go as big and are looking to save a few bucks, you can always shoot for the 3020 (which is also a more on-par comparison to the WhittleCNC and 3018 you linked to in terms of physical dimensions anyway).

thank you sir.  a thousand thanks.  that is a wealth of info and I very much appreciate it.  I was leaning towards the 3040 just because I was thinking it might allow me to do 'half templates' of a guitar with hopes of joining them together.  That might be 'pie in the sky'.

it gives me great confidence to hear you say you started out on something similar. 

I'm fine with the idea of a 1/2 mdf template taking a while... I'm thinking perhaps if I could play with one I could get comfortable enough to make big templates and pay someone with a bigger/faster machine to run them.  if I find I would use it a lot I think I can maybe convert this to lazer etch and buy something bigger.  I just don't want to spend $3k w/o ever having tried it as it might be a big waste. 

a thousand thanks for your reply.

[curtisa]

2 hours ago, mistermikev said:

I was leaning towards the 3040 just because I was thinking it might allow me to do 'half templates' of a guitar with hopes of joining them together.  That might be 'pie in the sky'.

By itself it won't be big enough to do most larger templates in one go, although with some careful planning you could do some 2x2 layouts and join all four quarters up to make a singular full size template of say, a body. I used to slot fretboards on the 3020 in two halves. Fiddly but do-able.

[mistermikev]

Just now, curtisa said:

By itself it won't be big enough to do most larger templates in one go, although with some careful planning you could do some 2x2 layouts and join all four quarters up to make a singular full size template of say, a body. I used to slot fretboards on the 3020 in two halves. Fiddly but do-able.

thank you again for the reply. 

i was thinking that the typical body is say 13.5 x 19... if the 3040 translates to 15.75 x 11.81, and assuming that's ACTUAL work space... I could do a body template in two 13.5 x 9.5 pieces giving me approx 1" of slop on all edges.  meaning bridge pickup|tremolo mount|butt of the guitar vs neck pickup|neck cavity|horns.  If they sort of had a finger joint between them... again perhaps this is pie in the sky but at least in theory I could make a 2 piece template for say an sg body.

fretboards - I imagine that would take a rock solid method of moving the workpiece away from the machine once the first cut is done.  I can't imagine the kind of atttention to detail that would take.  bravo.

[curtisa]

7 hours ago, mistermikev said:

was thinking that the typical body is say 13.5 x 19... if the 3040 translates to 15.75 x 11.81, and assuming that's ACTUAL work space... I could do a body template in two 13.5 x 9.5 pieces giving me approx 1" of slop on all edges

Ah, yes. I was thinking of splitting the body template up the centreline and orienting each long edge up the Y axis. But yes, you're right - nothing stopping you swinging the whole thing around by 90 degrees and using the longest axis of travel to cover the widest part of the body and doing a 1+1 arrangement instead.

7 hours ago, mistermikev said:

fretboards - I imagine that would take a rock solid method of moving the workpiece away from the machine once the first cut is done.  I can't imagine the kind of atttention to detail that would take.  bravo.

I wrote an article about the process some time ago describing how it can be done: Part 1 and Part 2.

Edit: ...whoops, and Part 3 as well.

 

[mistermikev]

I took a quick look at that... brilliant.  the bit themselves as the pin - epic.  One thing that kind of stood out - you commented your code!  love it. 

For now... it's a lot of 'above my head' but the nice thing about this thread is it just became a bookmark because I'm certain I will go back and digest that line by line. 

 

thank you so much for sharing that with pg and rubbing my nose in it here.  fantastic stuff.  I really appreciate you guidance!~  THANK YOU!

[mistermikev]

first off @MiKro I know you were working on a machine build and I feel guilty that I'm even considering spending near what you said you would be selling for... and I don't want to offend.  I'm sure the machine I'm contemplating is not nearly as good... but I'm fairly convinced I need a machine that there are lots of resources/threads/semi-support online... and that has pushed me to consider the 6040 - please forgive.

 

 

can anyone shed light:

this week I have been trying to read up on cnczone about dif machines.    from what I have read the 6040 machines are "much better built" (ball screws for all 3 axis, spindle that can do hardwood, and 5" x 15" x 25")  so I am contemplating this one:

https://www.ebay.com/itm/USB-4Axis-6040Z-CNC-Router-Engraving-Machine-Metal-Mill-Cutting-1-5KW-24000-rpm/392791430830

my expectations are fairly low.  If it can cut all the templates I need for a build, do pcbs... I'll be thrilled.

 

 

if I understand, the controller that comes with any of these models may be something I need to upgrade to a geko g540 at some point?  @curtisa did you observe this with yours?

I've started watching fusion 360 tutorials and have downloaded the free 1 yr version.  I'd like to stay free for now... and did not like freecad bugs.

I think I can use the free version of mach3 to run the machine and ensure it will work?  free version will only execute 50 lines so I won't be able to do much?  then another $175 to get a license? 

 

was hoping for a bit more guidance... are there any cliffs I'm about to go over?

 

[MiKro]

1 hour ago, mistermikev said:

first off @MiKro I know you were working on a machine build and I feel guilty that I'm even considering spending near what you said you would be selling for... and I don't want to offend.  I'm sure the machine I'm contemplating is not nearly as good... but I'm fairly convinced I need a machine that there are lots of resources/threads/semi-support online... and that has pushed me to consider the 6040 - please forgive.

>snip<

 

Mike I am not offended at all. I understand your wants. I will say that this is an environment that requires a lot of upfront study as well as trial and error, mix all of that with frustration.LOL

As far as the Chinese CNC's I will defer to others as I know little about them.

Mike

[MiKro]

2 hours ago, mistermikev said:

>snip<

I think I can use the free version of mach3 to run the machine and ensure it will work?  free version will only execute 50 lines so I won't be able to do much?  then another $175 to get a license? 

>snip<

 

 

Just an FYI the free version allows 500 lines which is nothing. A carved guitar body can use 1 mil or more lines

[mistermikev]

15 minutes ago, MiKro said:

Just an FYI the free version allows 500 lines which is nothing. A carved guitar body can use 1 mil or more lines

was thinking... if it's anything like vb or bash shell - I'll just pipe my arguments into more arguments and write one big line (just kidding, hehe). 

[mistermikev]

actually looking at it... bringing back flashbacks of machine code.  ewww!  never wanted to remember that!  bit translation - yuk!

[MiKro]

1 hour ago, mistermikev said:

was thinking... if it's anything like vb or bash shell - I'll just pipe my arguments into more arguments and write one big line (just kidding, hehe). 

Nope it does not work that way. I wish it did at times. it would be easier to write some things where the logic makes more sense. LOL

[mistermikev]

Just now, MiKro said:

Nope it does work that way. I wish it did at times. it would be easier to write some things where the logic makes more sense. LOL

yup, I  got that... was just being funny.  at a place where I worked we had a little joke code competition going on at one point... who could accomplish the most with just one line.  in theory you can go on forever in some languages but in practice it gets harder and harder to not ruin the whole line with a mistake that you then can't pinpoint.   good times!

[curtisa]

3 hours ago, mistermikev said:

from what I have read the 6040 machines are "much better built" (ball screws for all 3 axis, spindle that can do hardwood, and 5" x 15" x 25")  so I am contemplating this one:

https://www.ebay.com/itm/USB-4Axis-6040Z-CNC-Router-Engraving-Machine-Metal-Mill-Cutting-1-5KW-24000-rpm/392791430830

my expectations are fairly low.  If it can cut all the templates I need for a build, do pcbs... I'll be thrilled.

That's a pretty big step up from what you were considering earlier. Do you need the 4th rotary axis? I suppose you could use one to carve the back of a neck, but it's not a mandatory requirement and adds a lot more complexity to the mix (and an extra chunk of cash too). The recent changes to the personal-use version of Fusion 360 removes 4th axis support, so unless you want to pay for an additional subsciption to F360 you'd have to hunt for a different CAM software solution to use it.

The spindle will work fine for direct hardwood cutting, but that's taking things well beyond your initial scope of just doing PCBs and templates (if you can directly cut the wood, why would you make a template?). It's also a watercooled spindle, so you'd need to allow for a small pump, a sealed bucket of water and the hoses in your workshop.

Like the 3040 you looked at earlier, it's not clear what the interface is designed to work with. If it's GRBL there's plenty of free/low cost options to go with. If it's Mach you'll need to add the price of a license. As @MiKro suggests, 500 lines of code won't take you very far.

Not trying to push you away from the 6040. It's definitely a machine that would fit 90% of applications you'd encounter building a guitar (save for perhaps a neck-through instrument, a bass neck or really large body shapes), but there's no doubt that it's positioned as an advanced-beginner item.

 

4 hours ago, mistermikev said:

if I understand, the controller that comes with any of these models may be something I need to upgrade to a geko g540 at some point?  @curtisa did you observe this with yours?

I never found that with mine, but I never expected my unit to be churning out parts on an industrial scale either. It worked for what it was expected to do just fine.

Oh, and on the Fusion 360 changes again - the other thing Autodesk removed from the free version was rapid motions, limiting the fastest travel to the cutting speed. Cutting feedrate when the tool is engaged in the material might be 800mm per minute, but your rapid feedrate when the cutter is retracted from the work is defined by whatever limit you impose on your machine, which could be thousands of mm/min. On a small machine that's not going to translate to any real issue when in use. On a larger machine, if you need to quickly and frequently traverse across a long distance in between cutting motions (say, carving the back of a neck) that's going to add up to a significant chunk of extra time.