School me on the best machine for big hardwood

[bluesking]

I’ve spent a lot of time lurking on this great forum and planing my next machine. I’ve finally accepted the extent of my ignorance and am now totally in a state of decision paralysis - please help!!

I am an electric guitar luthier and for a few years I have used a small and crappy chinese 3018 CNC for inlay work and fretboard carving. Its been a trial but I have learnt a lot. I’m pretty skilled with using both handheld and tablemounted routers and working with wood in general. I also have a software and electronics background, reasonable mechanical savvy but little no metalworking experience.

I’ve decided to take the plunge into using CNC to carve full guitar bodies and necks. I will need a working area of at least 900 x 400 mm to achieve this - whilst bigger would certainly be more flexible (e.g. for occasional bass guitars necks and neck-throughs) I think keeping it smaller is better - for size and weight constraints. In the rare case I need to, I’m happy to tile the work.

Here are some of my initial thoughts which I’d really appreciate a sanity check on:

1) My feeling is that for reasonably large hardwood pieces I need to ensure a high feedrate is possible - at least 5000mm/min if not 10000. Accuracy is important but I think not as critical as for you metalworking guys - my feeling from my handbuilt guitars is that 0.1mm accuracy is totally sufficient. Is my thinking on the right track?

2) Rigidity obviously a key concern as it would be for any machine - I’d like to take as high a DOC as possible. I’ve seen lots of youtube videos of chaps CNCing guitars, and often I’m very surprised by the very small bits they seem to use (often around 1/4”). This seems really strange to me - I’d rarely use such a smal bit in a handheld router. The obvious thing is, I need a spindle capable of taking ER20 collets, but the more difficult problem is finding a machine/design that provides adequate rigidity to do justice to something like a 1” ball nose with a 3mm DOC.

3) Machine weight is a major headache. Its quite possible my workshop may need to relocate and I am concerned about having a 200kg behemoth to drag through doors and around stairs. I appreciate this is going to limit the rigidity I can expect! My best hope I think is having a machine that is straightforward to dismantle (ruling out the relatively attractive full table sized chinese offerings).

4) Cost is certainly a consideration but to some extent it is not the limiting factor - the weight problem dominates.

Here are my options as I see them:

1) Build my own. I have taken great interest in the build described here http://www.mycncuk.com/threads/11266-3-Axis-900x500mm. Similar requirements to my own and a largely modular aluminium design. Machined aluminium sideplates would be the biggest challenge for me but I can potentially access a CNC machine at my local hackspace. I’d look to use 4Nm Nema23 motors and add a pulley coupling mechanism - hopefully allowing me to tune the gearing for a good balance of accuracy/speed (also it seems much more difficult to get hold of 1610/2010 ballscrews compared to 1605 so I see gearing as a way around that). I could perhaps dedicate a full month of my time to this build and have costed it to around £3k - is this mad?

2) Omio 9060 CNC for around £3k all in (https://www.omiocnc.com/x8-2200epl.html). Looks like a generic chinese machine but with all the trimings (2.2kw water cooled spingle, square supported rail, ballscrews, bellows, reasonably quick shipping). Would this be much worse than option 1? In some ways the basic design is very similar to the design for 1, aluminium extrusions with aluminium endplates. Motors are directly coupled though and a max speed of 4000mm/min which seems a little slow. A respectable performance in the following video even using a tiny router cutter for some reason (https://www.youtube.com/watch?v=HyMO89obfmg)

3) A brace of comparatively heavy and rigid machines which utilise at least some steel frame components. These tend to be more expensive and not to use square supported rail and some even use ACME lead screws - on paper this sounds like bad news but I wonder whether the overall machine rigidity is more important than the “sum of the parts”?

a) JBEC 106512
£7k+ and a lukewarm mycncuk reception (http://www.mycncuk.com/threads/6389-JBEC-CNC-Router-System-is-now-available-in-the-UK)
but I have seen some videos of quite impressive guitar building performance (https://www.youtube.com/watch?v=OCC_4lJcRAI)

b) Kompas H - 1000 GS KIT
Eastern european wildcard, £1.5k for a mechanical only soltuion - could do the electronics myself but max speed seems really low

c) CRP4828
Mechanical only for £2.5k - could do the electronics myself. Seems like old technology, using roller bearings and ACME leadscrews. Still has admirable performance as in this example (https://www.youtube.com/watch?v=1VEebB157G4). Maybe the rigid frame and modular construction make it a winner?

d) I tech K6090T (https://www.scosarg.com/itech-k6090t-q-series-desktop-cnc-router)
£5.5k and seems to have all the right qualities - square rails, ballscrews, fast traverse! This seems to be the pick of the bunch but at 200kg I worry about transportability. How would you move this around if it was yours? Theres an unboxing video of this machine on youtube and they use a chain hoist to get it up on the workbench!

Wow, that was quite a lot. If you got this far I’m really grateful. Any help or advice you can give would mean the world to me at this stage!

[Washout]

Hi bluesking

I also machine guitars on my CNC machine (check my signature for my channel). I normally machine around the 3000mm/min mark using 4 to 10mm cutters dependant on operation and strategy (e.g. larger multi-flute cutters are great for adaptive/trochoidal clearing when roughing bodies and body pockets).

Personally I would go for option 1. as you will know what you are getting and can design out some of the compromises in the other options. £3K is not unreasonable for a decent aluminium framed machine (mine was about £2K-£2.5K including software).

Option 2 looks OK, but I'd like to see how that X axis (long axis) is being driven and the electronics might be a bit of a lottery (i.e. stepper type/voltage, driver type, controller board etc.).

Option 3 a. not familiar, but if there are valid criticisms from JAZZCNC or the other regulars here then I would heed them.
Option 3 b. not familiar so can't give an opinion
Option 3 c. I would not go for any machine with leadscrews and probably not without linear guides/bearing (square supported rails) given the choice these days.
Option 3d. You will likely find that many machines will start to get close or pass the 200Kg mark, once you get it fully loaded - generally provided the motors are sized correctly (and some other factors), heavier is better

Hope that helps.

[bluesking]

Hi bluesking

I also machine guitars on my CNC machine (check my signature for my channel). I normally machine around the 3000mm/min mark using 4 to 10mm cutters dependant on operation and strategy (e.g. larger multi-flute cutters are great for adaptive/trochoidal clearing when roughing bodies and body pockets).

Personally I would go for option 1. as you will know what you are getting and can design out some of the compromises in the other options. £3K is not unreasonable for a decent aluminium framed machine (mine was about £2K-£2.5K including software).

Option 2 looks OK, but I'd like to see how that X axis (long axis) is being driven and the electronics might be a bit of a lottery (i.e. stepper type/voltage, driver type, controller board etc.).

Option 3 a. not familiar, but if there are valid criticisms from JAZZCNC or the other regulars here then I would heed them.
Option 3 b. not familiar so can't give an opinion
Option 3 c. I would not go for any machine with leadscrews and probably not without linear guides/bearing (square supported rails) given the choice these days.
Option 3d. You will likely find that many machines will start to get close or pass the 200Kg mark, once you get it fully loaded - generally provided the motors are sized correctly (and some other factors), heavier is better

Hope that helps.

Washout, that really does help, thank you very much!

Option 1 certainly is attractive but I am somewhat concerned about disappearing down the rabbit hole into a long build. I'm prone to this anyway, and I can see myself getting carried away! At least if it turns out heavy I know I can easily dismantle it for transportation which solves that problem.

Option 2: X axis is driven by a single, central 2005 ball screw. As for electronics, they seem OK on paper. For my own build I was thinking about using this (https://www.cnc4you.co.uk/CNC-Kits/Professional-Stepper-Kits/Stepper-Motor-PRO-Digital-Kit-4Nm-x-3-Axis-Nema-23). The Omio seems similar to what I would have used in my build, I think the drivers will be about the same, the PSU will run a little lower at 40V rather than 48V, and the steppers seem like they are 3Nm rather 4Nm (can't really tell, only info on the steppers is they are rated 4.2A - no idea what torque this corresponds to). Not really sure how different the BOBs would be. I don't know whether this all would really impact what I'm doing as my own design was based on what can easily be bought in at low risk (given they are packaged as a kit I was relying on compatibility), rather than on any real engineering design factors. Perhaps this kit would be overkill for what I need and perhaps the Omio would be just fine - or perhaps not?!?! Overall, at 95kg and with an easily removable gantry I would feel comfortable knowing I can also move it around.

The biggest differences I can see between the omio and my own design are:

a) Omio uses 5mm lead pitch with direct coupling between steppers and leadscrews - I was going to design belt-gearing to get a bit of extra speed. Either should be able to run at over 3000mm/min so perhaps this is irrelevant, given what speeds you have found work best. I'm hoping to use more standard router cutters than milling bits, 1/2" flat bits for clearing, 1" ballnose for contours and top carves - not really sure if these expectations are overambitious or just plain crazy - not seeing anyone else doing this - wonder why? Those are the bits I'd use when hand routing and I can't say the forces involved seem dramatic at all - some of my handbuilt guits have made use of these bits supported by pretty rudimentary MDF sleds and jigs - sure there is often around 0.5mm of flex - never caused any problems after sanding and finishing! Hell, I'm currently using an angle grinder for most of my top-carves - not fun, not accurate, but gets a great result in the end.

b) I would have used significantly larger extrusions - its hard to tell what size extrusions the omio utilises - but then the omio makes more use of more aluminium plate - have no idea what rigidity to expect from either option!

c) Omio will hopefully be much easier and quicker to get going on. Also, the bellows seem a nice touch, I have the feeling that the omio will be more protected from sawdust than anything I will ever end up designing.

I have just had a very reasonable price from Omio for supplying the X8 without the electronics or steppers. So a hybrid approach may be best (and even cheapest). With this I'd probably end up combining the omio mechanics with the cnc4you electronics kit and.... hoping.... that Robert really is your father's brother?!?!

I had a look at your youtube channel and realised I have already seen all of your Andy Summers tele build
during my research. I think I've seen every guitar CNC video on youtube by now haha. Really enjoyed your work, particularly nice and sharp sunburst you got on that one!

Cheers!

[Clive S]

First welcome to the forum bluesking.

I would also go with option one as then you are in control. Have a look at Joe Harris build with a hand held router https://www.youtube.com/watch?v=pZo2LVQA9UI&list=PL1FIADAKba_uTgFU5qqS3i705fuSogBXT

The motor kit is not good as the drives and power supply are not good enough. cnc4you is a good place to buy the motors (or Zapp. https://www.zappautomation.co.uk/electrical-products/stepper-systems/stepper-motors.html)
Wal's machine is currently being modified to make it into a twin screw type. I would advise you to do the same. This can be done with two screws and two motors nema 23 OR two screws and one motor nema 34 with one belt connection both screws and the motor. Both ways are fine personally I prefer two motors. The are pros and cons for both types.

I would go with about 68V power supply and em806 drives or am882 or even https://www.ebay.co.uk/itm/CNC-2DM860H-Digital-Microstep-Driver-Stepper-Motor-Controller-32bit-DSP-Blue/293026216828?hash=item4439b93b7c:g:31oAAOSwKWJcnaQ E.

Ball screws 1610 x and y 1605 on z tried and tested by many.

If you decide to build start a build log to keep all the question in one place. Good Luck

[bluesking]

First welcome to the forum bluesking.

I would also go with option one as then you are in control. Have a look at Joe Harris build with a hand held router https://www.youtube.com/watch?v=pZo2LVQA9UI&list=PL1FIADAKba_uTgFU5qqS3i705fuSogBXT

The motor kit is not good as the drives and power supply are not good enough. cnc4you is a good place to buy the motors (or Zapp. https://www.zappautomation.co.uk/electrical-products/stepper-systems/stepper-motors.html)
Wal's machine is currently being modified to make it into a twin screw type. I would advise you to do the same. This can be done with two screws and two motors nema 23 OR two screws and one motor nema 34 with one belt connection both screws and the motor. Both ways are fine personally I prefer two motors. The are pros and cons for both types.

I would go with about 68V power supply and em806 drives or am882 or even https://www.ebay.co.uk/itm/CNC-2DM860H-Digital-Microstep-Driver-Stepper-Motor-Controller-32bit-DSP-Blue/293026216828?hash=item4439b93b7c:g:31oAAOSwKWJcnaQ E.

Ball screws 1610 x and y 1605 on z tried and tested by many.

If you decide to build start a build log to keep all the question in one place. Good Luck

Thanks Clive,

Noted your advice about the drives and power supply - I had a nagging feeling this was a bad area to scrimp on but my choice was driven by ignorance rather than price. I'm sure if I went with your suggestions I'd be able to find someone here who has a wiring diagram for all these exotic options (exotic to me anyway!)

Also noted the need for substantial motoring to move the gantry.

Joe Harris build video is interesting. It looks like a sturdy machine, but what do I know?! It wouldn't work for me - as he rightly points out, making a steel frame structure which is easy to dismantle and reassemble without fubaring its alignment seems like a no go - I am currently moving workshop space and it seems quite likely I will have to move again within the foreseeable future - I can't afford a machine which I can't easily take with me!

I have already started design work in CAD for my build (2 variants actually). Once I have something ready to show I will definitely start a build log and seek comments.

To press on with my design I now really need to commit to either a moving gantry or a fixed gantry. I originally went with a fixed gantry based on what I saw this Maltese luthier doing (https://www.youtube.com/watch?v=6MCyIZIcg48). Its a bit rough and ready looking, but my mechanical intuition is telling me it is a good way forward. Certainly moving 2kg of wood seems like a much easier job than an entire gantry holding a 2.2kw ER20 spindle. Then I saw the thread of Wal's machine and I liked the simplicity and elegance of it - so I went down that route. Now I'm dithering again between the two (can you guess I'm a ditherer yet, hahaha?)

[Clive S]

To press on with my design I now really need to commit to either a moving gantry or a fixed gantry. I originally went with a fixed gantry based on what I saw this Maltese luthier doing (https://www.youtube.com/watch?v=6MCyIZIcg48). Its a bit rough and ready looking, but my mechanical intuition is telling me it is a good way forward. Certainly moving 2kg of wood seems like a much easier job than an entire gantry holding a 2.2kw ER20 spindle. Then I saw the thread of Wal's machine and I liked the simplicity and elegance of it - so I went down that route. Now I'm dithering again between the two (can you guess I'm a ditherer yet, hahaha?)

Fixed gantry will take up more floor space and moving gantry. if you are comparing that with the one Wal built then there is no comparison between the two. I pointed you to Joe's machine not for the way it is built but the methods he used with the hand held router. Also he has on his site a very good vid on building the power supply. There are only three components in it.

I forgot but go for 20mm square rails if you can. all this can be got from Fred at BST in China along with the ball screws and bk bf bearings etc.


Also noted the need for substantial motoring to move the gantry. nemas 23 will be more than enough with the correct drives.

Re the CAD it is always a good idea to do a design first. I don't think Wal's machine is anything like 100Kg

[bluesking]

Thanks again Clive,

Something has been sitting uneasy with me about all of this and I think I may have finally figured it out. When I'm seeing various machines cutting guitars on Youtube I'm generally left underwhelmed. I think it is most likely that I come from a purely woodworking background and metalwork is a foreign world to me. But what I actually need is a woodworking machine, I couldn't care less about aluminium (above and beyond any I need to machine to produce the CNC itself).

The only truly satisfactory results I've seen to date have been on HAAS-type machines. I've obviously been managing my expectations and ruled out anything in that league. And then suddenly I see this:

https://www.youtube.com/watch?v=LLNeA1oUCGc

This is the first time I've seen a desktop machine doing exactly what I hoped I could do. So I start trying to pick it apart - why is this machine doing what I require while I see no evidence that others are? Hard to see on the video, and the manufacturer K2CNC is now kaput. I follow some breadcrumbs and find some images of the machine:

https://www.hmres.com/site/images/productimages/2db2ca44-1d3d-40e5-8e6f-87f6.jpg

Looks OK, but pretty unremarkable. Yes its got square rails, yes, its probably using ball screws, yes it doesn't use steel, yes, its only using a single screw for the long axis, yes, it doesn't look all that sturdy. The only two remarkable things I notice are:

1 - Monster spindle - I don't know what power, but looks to be at least 3kW
2 - SERVOS!!!!

You can probably guess where my mind is going - and it does relate to my earlier whinging about everyone else using what I consider to be "small" cutters. I really want to use the cutters that I understand, and these are mostly big, multi-flute affairs. Now, my understanding is that I need to have substantial feed rates to generate good chips in such situations. My early research suggests steppers can certainly yield the same torque as servos, but also that servo's reach peak torque at high speeds, while steppers reach peak torque at low/zero speeds. Servos match my personal experience of hand-routing wood - a certain minimum feed rate is necessary to achieve controllable and clean cuts.

I get the impression that servos are a little rarer in most builds here and I understand they bring lots of extra cost and complexity. I'm particularly concerned that if I go down the servo route I may be out in no-man's land and perhaps less able to rely on the expertise of people here. Is there anyone here who can hold my hand a little and give me some advice about servos?

I may well be wrong, maybe steppers are perfectly sufficient, but I can now only convince myself of this if I can see them producing similar results to the above youtube link. At the end of the day, if I can't expect that sort of performance, I don't think I can justify investing so much time in a custom build!

[Clive S]

I may well be wrong, maybe steppers are perfectly sufficient, but I can now only convince myself of this if I can see them producing similar results to the above youtube link. At the end of the day, if I can't expect that sort of performance, I don't think I can justify investing so much time in a custom build!

Yes bigger more powerful spindles like 3 kw you can use bigger cutters. If you want more motor power then use nema 34 but then you need higher voltage drives (mains voltage type).

The machine like Wal's in my opinion is far more sturdy than the one you linked to. Modern tool paths can out perform with greater speed and less spindle power. Why not have a look at some feed and speed calculators.

using 10mm pitch screws with steppers you can easily get 10mts/min You will be surprised how fast 5mts/min is on a small machine.

[Voicecoil]

Your requirements are not a million miles away from what I want to end up with, except mine will be used for building high-end loudspeakers rather than guitars (hence needs to be a bit larger); I've also had to think about the relocation issue. My advice if you're going to build your own is to go for a "raised side rail" arrangement, as then it's pretty easy to make the gantry detachable which would obviously help with moving the thing. But even so, I wouldn't be expecting the base part of anything that's properly rigid to be a one man lift - however if you're making one yourself you can of course build in handles/lifting points and provision for mounting it on a dolly or similar. I've also had a word with my pal Mike Vanden, one of the best archtop luthiers in the country, and he said don't get too hung up on feed rates as with some fancy hardwoods you can get a poor cut if you go too fast - he tends to work his Isel machine at 2000mm/min tops on standard wood and less if doing fine work on brittle stuff. A lot of the problem is that wood isn't a completely even material and some hardwoods are a bit on the brittle side; if you hit a hard spot/twisty bit near a corner on the end grain you can easily take a wee chip out of the edge. Using downward spiral cutters and machining the end grain first can help obviously, but won't solve everything all the time.
Regarding your comments on the K2CNC machine, it's difficult to say exactly what size that spindle is, but assuming the square rails are 20mm, I guess it could be a 2.2KW aircooled thing. Although servos have better torque at high rpm than steppers that doesn't mean you can't make a machine that goes fast enough for your kind of work using steppers, it's just a matter of getting the design right, there's a very useful motor/ballscrew calculator on this forum. And then there's now the closed-loop steppers available which are kind of a half-way house.

[bluesking]

Yes bigger more powerful spindles like 3 kw you can use bigger cutters. If you want more motor power then use nema 34 but then you need higher voltage drives (mains voltage type).

The machine like Wal's in my opinion is far more sturdy than the one you linked to. Modern tool paths can out perform with greater speed and less spindle power. Why not have a look at some feed and speed calculators.

using 10mm pitch screws with steppers you can easily get 10mts/min You will be surprised how fast 5mts/min is on a small machine.

Thanks Clive, good food for thought.

Wal's machine certainly looks more sturdy than this K2. In fact, almost all machines I've seen look more sturdy than the K2. In my mind, this begs the question, why have I not seen any desktop machine performing as well as the K2 in this scenario?

It doesn't look like the timings will work out for me to see Wal's machine in action before pressing ahead with mine as I do need to get going. I'm sure you're right about 10/5 m/min, I couldn't even tell you what numbers the K2 is running at - all I can see is that the cuts look exactly what I want to be doing and I've not seen anything else that doest - leaves me with quite the risk of going down a dark path!

Compare the rickety K2 taking nice deep cuts with big cutters:
https://www.youtube.com/watch?v=LLNeA1oUCGc

To these much more sturdy machines, seemingly taking their sweet time scratching away:
https://www.youtube.com/watch?v=F5Jy5cJscXU
https://www.youtube.com/watch?v=WatC-mihuoA
https://www.youtube.com/watch?v=Psr9Yt49grw

[bluesking]

Hi Voicecoil, thanks for your help.


My advice if you're going to build your own is to go for a "raised side rail" arrangement, as then it's pretty easy to make the gantry detachable which would obviously help with moving the thing.

By "raised side rail" Im guessing you mean something like Washout's machine (https://www.youtube.com/user/CCWashout/videos). Doing this seems to trade gantry weight for base weight, which sounds like a good idea for general performance but doesn't that just put more weight on the already heaviest section when it comes to transportation?



But even so, I wouldn't be expecting the base part of anything that's properly rigid to be a one man lift - however if you're making one yourself you can of course build in handles/lifting points and provision for mounting it on a dolly or similar.


I needed to hear this, even if I didn't want to! My best hope I think is to rely on modular design so I can dismantle the base into components for transportation purposes. At least then I know I won't find myself stuck somewhere.



I've also had a word with my pal Mike Vanden, one of the best archtop luthiers in the country, and he said don't get too hung up on feed rates as with some fancy hardwoods you can get a poor cut if you go too fast - he tends to work his Isel machine at 2000mm/min tops on standard wood and less if doing fine work on brittle stuff. A lot of the problem is that wood isn't a completely even material and some hardwoods are a bit on the brittle side; if you hit a hard spot/twisty bit near a corner on the end grain you can easily take a wee chip out of the edge. Using downward spiral cutters and machining the end grain first can help obviously, but won't solve everything all the time.


I appreciate that advice. I am not an arch top builder - just carved tops for me. I use everything from angle grinders to French curve scrapers to fashion my tops at the moment - inconsistency in wood density and fibre is par for the course. I can't speak for Mike Vanden, only for myself, nor can I put numbers to any of my naive expectations, but I will continue to assert that I do want this performance for my top carving: https://www.youtube.com/watch?v=LLNeA1oUCGc



Although servos have better torque at high rpm than steppers that doesn't mean you can't make a machine that goes fast enough for your kind of work using steppers, it's just a matter of getting the design right, there's a very useful motor/ballscrew calculator on this forum. And then there's now the closed-loop steppers available which are kind of a half-way house.

Sounds reasonable. I think I need to spend some time comparing torque curves. Whilst I'm comfortable with HT voltages from my valve amp work, Clive's talk of NEMA 34s and HT voltages is freaking me out from a complexity point of view.

I've just read about ClearPath servos, is that what you mean by closed-loop? I'm starting to understand there are a few more shades of grey between stepper and servo than I'd initially thought.

[Washout]

Hi bluesking

I can see why you want to go for larger cutters and for roughing I would agree, however there is another factor to maybe consider dependant on the guitar design/type you are looking to machine. I'm sure you're aware that most routed pockets on say a Les Paul or Telecaster are a. imperial, but b. also are designed for a set of cutters that have certain corner radii e.g. 1/8 inch which means using a 1/4 inch cutter for most humbuckers. So you may be able to rough out the pockets with an 8-12mm cutter, but will have to do a tool change and a clean up operation with a smaller cutter to clear out the corners.

Also I am using 70v nema 23 steppers on my machine and it easily goes to 15mm depth per pass and I could make that 20mm without too much if any additional tool wear - see here: https://youtu.be/WVf5J8XytSA

BTW if you want the CAD for a "normal" Tele I have that linked in video #9 of the series above. There's a thread on here somewhere as well, as it was a kind of community project.

Chris

[bluesking]

Hi bluesking

I can see why you want to go for larger cutters and for roughing I would agree, however there is another factor to maybe consider dependant on the guitar design/type you are looking to machine. I'm sure you're aware that most routed pockets on say a Les Paul or Telecaster are a. imperial, but b. also are designed for a set of cutters that have certain corner radii e.g. 1/8 inch which means using a 1/4 inch cutter for most humbuckers. So you may be able to rough out the pockets with an 8-12mm cutter, but will have to do a tool change and a clean up operation with a smaller cutter to clear out the corners.

Also I am using 70v nema 23 steppers on my machine and it easily goes to 15mm depth per pass and I could make that 20mm without too much if any additional tool wear - see here: https://youtu.be/WVf5J8XytSA

BTW if you want the CAD for a "normal" Tele I have that linked in video #9 of the series above. There's a thread on here somewhere as well, as it was a kind of community project.

Chris

Hi Chris,
In my experience, traditional LP and tele style guitars are designed to be cut using 1/2" diameter cutters - you can cut pickup cavities, neck pockets and control cavities all using a 1/2". This is also true for a basic Strat design too. It is certainly true for my own designs, which are my focus.

Good to know what sort of cutting performance you are achieving, the results look good. My biggest focus is on top carving - in fact I only went down the CNC path after deciding that I would build a duplicarver for this job alone - and realised that it would be crazy not to look into CNC. Do you have any example of top carves you've produced on your machine?

Thanks for the CAD link, can never have too much source material.

[Washout]

I don't have any drop/carve tops yet, but do have the design run up for an RG/Super Strat custom design with a carve which I'll be cutting, once I have finally finished my AS Tele project. The strategy I'll be using for that will be similar to the video you linked, but will use Fusion 360's Adaptive (trochoidal) clearing rather than a traditional slot/side milled waterline strategy. The reasons for that are that it vastly reduces tool and machine wear and you can run at very high speeds (with careful tuning) - you can see the strategy at work on that Tele video when its roughing out the front side cavities where I'm running at approx 3,500mm/min on a single flute cutter, but could increase that dramatically using 2, 3 or 4 flute cutters.

The carve will then be "smoothed" out on a finishing pass using a ballnose cutter of 8+mm, which should then only leave light sanding. You can see some of that in the neck cutting videos on my channel.

The "problem child" cuts for guitars are the final cut outs I've found, where unless you have the room for "adaptive clearing" in a gutter around the body shape, you'll be using slot milling, which means reducing the depth of cut, as it tends to brutalise the machine and cutter otherwise. Doesn't mean it won't do it, but it just takes a lot longer than adaptive clearing. Again you can see the slot milling cut near the end of that video to compare feeds and speeds.

Hope that helps.

[bluesking]

I don't have any drop/carve tops yet, but do have the design run up for an RG/Super Strat custom design with a carve which I'll be cutting, once I have finally finished my AS Tele project. The strategy I'll be using for that will be similar to the video you linked, but will use Fusion 360's Adaptive (trochoidal) clearing rather than a traditional slot/side milled waterline strategy. The reasons for that are that it vastly reduces tool and machine wear and you can run at very high speeds (with careful tuning) - you can see the strategy at work on that Tele video when its roughing out the front side cavities where I'm running at approx 3,500mm/min on a single flute cutter, but could increase that dramatically using 2, 3 or 4 flute cutters.

The carve will then be "smoothed" out on a finishing pass using a ballnose cutter of 8+mm, which should then only leave light sanding. You can see some of that in the neck cutting videos on my channel.

The "problem child" cuts for guitars are the final cut outs I've found, where unless you have the room for "adaptive clearing" in a gutter around the body shape, you'll be using slot milling, which means reducing the depth of cut, as it tends to brutalise the machine and cutter otherwise. Doesn't mean it won't do it, but it just takes a lot longer than adaptive clearing. Again you can see the slot milling cut near the end of that video to compare feeds and speeds.

Hope that helps.

Thanks Chris,
A little over my head! I'm certainly no CNC expert, my work to date has stopped at the CAD point (using CAM only for inlays and fret slots). I know I'm probably putting the cart before the horse but I need a machine that will be capable of working in a way I understand - I've clearly got a lot of learning ahead of me.

Worst case scenario, if I can at least automate template production I will already be way ahead. My latest design takes me a couple of weeks to manufacture by hand - I'm not expecting it to jump off the CNC ready for finishing:

25688
25689

If I can cut the time down to 1 week I will be happy, but I need a machine that will bend to my needs rather than vice versa.

[Voicecoil]

Hi Voicecoil, thanks for your help.


By "raised side rail" Im guessing you mean something like Washout's machine (https://www.youtube.com/user/CCWashout/videos). Doing this seems to trade gantry weight for base weight, which sounds like a good idea for general performance but doesn't that just put more weight on the already heaviest section when it comes to transportation?

Yes, something like that, though if it were me I'd have tall side rails fixed to the base all the way along rather than perched on blocks - that way they add good longitudinal strength which will win you back a fair bit of the extra weight.Even without taking account of that they need only add 10Kg or so. And the gantry won't be overly light either - with a decent spindle and z-axis, around 30Kg or more is likely to be where it's at.


I needed to hear this, even if I didn't want to! My best hope I think is to rely on modular design so I can dismantle the base into components for transportation purposes. At least then I know I won't find myself stuck somewhere.


I reckon you could get the base part (assuming it''s a table-top thing) to be a 2 man lift - all you need then is a fit mate to help you for an hour or 2. And like I said before build provision for some handles into the design then it will be A LOT easier to move


I've just read about ClearPath servos, is that what you mean by closed-loop? I'm starting to understand there are a few more shades of grey between stepper and servo than I'd initially thought.

It wasn't the Clearpath units I was meaning, but they are another shade of grey - and likely easier to use than a classic servo from what I've just read. Closed loop steppers are a stepper motor with an encoder on the back which gives feedback to make sure it got to the target position - check out brands like Leadshine and Lichuan.

[bluesking]

It wasn't the Clearpath units I was meaning, but they are another shade of grey - and likely easier to use than a classic servo from what I've just read. Closed loop steppers are a stepper motor with an encoder on the back which gives feedback to make sure it got to the target position - check out brands like Leadshine and Lichuan.

I think I understand. Something like this?
https://www.cnc4you.co.uk/Stepper-Motor-Plus/Stepper-Servo-Motor/Stepper-Servo-Kit-4Nm

Clearpath do seem attractive as they are proper servos which should easily drop in place of a traditional stepper. One caveat seems to be that they are not well suited for use where multiple motors serve a single axis - though they do provide closed loop control, this control loop never exits the motor, so synchronising multiple motors seems impossible/difficult.

[Voicecoil]

Aye, they're the things - a lot cheaper on Aliexpress though! When I get my machine finished I'll bum a bit of hardwood off Mike and see how smoothly they cut it.Theoretically I can't see that there would be a lot of difference between a Clearpath and a closed loop stepper, but as always the devil will be in the detail which in this case will be the algorithms in the driver modules.

[bluesking]

Aye, they're the things - a lot cheaper on Aliexpress though!
Indeed, I'm trying to plan my build around locally sourceable components. I'll buy them from china when I'm ready but I want to know that I can fallback to something local if it gets me out of a jam or around a big delay.


When I get my machine finished I'll bum a bit of hardwood off Mike and see how smoothly they cut it.Theoretically I can't see that there would be a lot of difference between a Clearpath and a closed loop stepper, but as always the devil will be in the detail which in this case will be the algorithms in the driver modules.

I would appreciate that hugely. I assume your build is using closed loop steppers?

I'm seeing the following shades of grey at the moment:
1) open loop steppers
2) closed loop steppers
3) closed loop steppers with integrated control (e.g. https://www.zappautomation.co.uk/electrical-products/integrated-motors/integrated-easy-servo-motors.html)
4) servos with integrated control (e.g. clear path or something like this https://www.zappautomation.co.uk/electrical-products/servo-systems/integrated-servo-motors/syint080-750w-integrated-dc-servo-motor.html)
5) traditional servos (controlled from a controller)

Here is a very useful thread on this forum - clarified a lot about servos for me:
http://www.mycncuk.com/threads/11096-Advice-on-buying-servos

[bluesking]

I've been hammering the research trying to find a reasonably simple and economical way of accessing K2 servo performance. I've managed to gather as much data as I can on the electronics in that K2 system to copy it: https://www.machsupport.com/forum/index.php?topic=462.0 (notice this post, although over 10 years old, is by yet another luthier, hint hint, nudge nudge)

Promising candidates were closed loop steppers, I think they probably offer huge advantages but I'm left unsure as to how close they would perform to the K2 I've become so enamoured with.

Clearpath servos look very simple to integrate, but if you want to buy into their ecosystem things get expensive real quick and their torque performance curves leave me questioning how closely their performance mirrors a traditional servo. (https://www.teknic.com/clearpath-accessories-guide/accessories-guide-sd-dc/#overview)

DMM kits also look good, but not cheap and it is hard to spec them for similar performance (https://store.dmm-tech.com/products/3-axis-0-4kw-dyn2-ac-servo-system)

Chinese servo/driver offerings are well priced, but scarcity of documentation, as well as the rareness of servo builds around here makes me question my ability to integrate them. (https://bstmotion.aliexpress.com/store/group/Servo-Motor-Driver/314742_256838494.html?spm=a2g1y.12024536.pcShopHea d_324926.1_6_2)

Many of the above use non-standard mounts (even if they do claim to be NEMA23 compatible!) which is a ball ache for someone who just wants to get the job done.

Finally, I followed breadcrumbs from the Gecko G320X drivers used on the K2 and found this. I think its quite reasonably priced, specced similar to the K2 (just a little bit better all round) and comes as a kit, with a decent wiring diagram, and using fairly common (if American-centric) components:

https://www.automationtechnologiesinc.com/products-page/cnc-3_axis_4_axis_kits/3-axis-nema23-350ozin-60v20a-psu-g230x-gecko-driver/

Theres also NEMA 34 versions, and an interesting kit with 2xNEMA23 1xNEMA34 axes:

https://www.automationtechnologiesinc.com/products-page/cnc-3_axis_4_axis_kits/3-axis-nema34/

Conveniently, the drawings show that the NEMA23 mounts are indeed standard and that the shafts are 1/4" which will make it easier to buy-in compatible pulleys for the inevitable gearing reduction.

I'd welcome any comments on these components as they are my current top pick for my machine.

One of their downsides is that I think brushed DC servos are used - so they will have a finite life, I will swallow that.

Another downside is they have no auto-tune capability, so I'll have to fiddle with trimpots and microswitches to tune them for my purposes. Luckily, I'll be coming at them armed with a few control systems courses under my belt and an oscilloscope!

[Voicecoil]

One of their downsides is that I think brushed DC servos are used - so they will have a finite life, I will swallow that
At least the brushes appear to be easily externally accessible - so easy to change I would guess.

Do those motors have integral brakes? If not, you might need to add one on the Z axis if you're going to run a big fat (end hence heavy:rolleyes: ) spindle - servos don't have the stiffness that steppers have when the power's off, and depending on gearing etc., you might get a slowly drooping spindle when you switch off.

[bluesking]

Do those motors have integral brakes?


An excellent point which I had not considered! Unfortunately no brake.

I plan to use the following 3.5kW spindle and VFD:
https://www.aliexpress.com/store/product/CNC-Engraving-Milling-Grinding-Ceramic-Bearing-Electric-Square-Wings-Air-Spindle-Motor-3-5KW-Fan-Cooling/807755_32801191438.html?spm=2114.12010612.8148356. 16.1bdf23a8f7ed5m

https://www.aliexpress.com/store/product/CNC-SPINDLE-VFD-DZB312B003-7L2DK-ORIGINAL-FULING-INVERTER-3PH-3-7KW-220-NEW-FOR-ENGRAVING-MILLING/807755_32311395333.html?spm=2114.12010612.8148356. 5.2e6c275cpko9w1

At 8kg, this spindle may well cause problems.

I will most likely use a fairly high gear reduction on the Z (maybe 5:1 or even higher as compared to 2:1 on X/Y). Perhaps this will amplify any holding force/friction enough to keep it in place when powered off. If it does turn out to be a problem I guess I could always buy an external brake like this (though I won't have a convenient 24v supply for it, nor will I know how to connect it, maybe using a relay on the BOB or something...)
https://www.zappautomation.co.uk/24v-holding-brake-2nm.html

I'm also not particularly happy about the C10/C11 breakout board that comes with the kit. The wiring diagram is a bit scant and I am not sure it has all the features I need for spindle control - and I don't know what a charge pump is, or if I need one, but it doesn't seem to have one.

If I am feeling up to it I may use the kit as an inspiration to piece together my own system using some more custom parts - but I'm already overwhelmed by the complexity of all this!



you might get a slowly drooping spindle when you switch off.

That's what she said

[Voicecoil]

FY! a charge pump is an electronic circuit that's usually used to generate a low-ish power but higher DC voltage from the existing DC supply using capacitors, rather than an inverter/transformer/boost converter arrangement - e.g. generating 10V from a +5V supply.


That's what she said

lol

[bluesking]

I've created a build log where I can press ahead with my ideas: would welcome any more comments you guys may have!