Calling All CNC Peoples: Building Your Own Group

[malcolmputer]

Hello @Team_Machine_Shop and @CalvinStence

My time in quarantine, combined with the potential reopening of DMS has sparked an idea. I don’t know about the rest of you, but at this point, I’ve almost forgotten how to run the HAAS and I don’t have any way to keep those skills sharp at home. While I do have a 3D printer, I’m woefully under prepared for machining at home and I’d like to kill both birds with the same stone.

With that in mind, I was considering building my own 3 axis CNC for home use. I’d like something that can be reasonably moved around, and not take up a whole room, but still be functional enough to so some light machining in metal (even if only aluminum).

I’d like this to be a collaborative project, with the project trying to get as many different backgrounds of people attracted to it, so we can get a maximum amount of learning from it, I’m open to ideas about the flow of the “classes” or “meetings” but my thought was a couple design meetings, some CAD/CAM work days, then some machining parts days, and finally some meetup and assembly/test operations days.

With the above in mind, I’m doing a call for research and ideas about a CNC machine that is:

• Open Source / Free Designs
• Small enough to be reasonably moved by one person and fit in a passenger vehicle. Ideally “3D printer sized”
• Can machine at minimum Aluminum, with at least enough rigidity to make a 0.010" depth of cut (so no Tormachs lol @michaelb )
• Must be primarily manufacture-able with the tools we have a DMS. Obviously, we’ll buy some servos/steppers and the linear rails etc, but the major chassis parts and bits and bobs should be machinable on our HAAS and lathes/mills
• Must not cost an arm and a leg. I’m flexible on this one, but designs over $1k will probably be a bad idea for group size, I was thinking of something in the sub $800 range. If the group is large enough, a group buy of materials will probably help this a bit.
• Have at least a 6" by 6" by 6" working area. We may be flexible on this, but we’d need to vote on it in the group.

• I’m In! Sounds fun!
• It’s a Pointless Dumb Idea
• Too much Money
• Too little build size
• Other

[CalvinStence]

It’s a bit over 1k… just a bit

[TBJK]

@CalvinStence

It’s a bit over 1k… just a bit

So 1,001.00. Lol

[frank_lima]

If you want a quality low cost CNC router check out the Omio X8:
https://www.omiocnc.com/products/x8-2200l-usb.html

I know quite a few people in the robotics community that have and swear by them. Here are some videos of the Omio cutting aluminum:

[tmc4242]

$2800 is out of range.

Looks nice, but too high for me at least.

[Kentamanos]

My current plan is to build one of these soonish™: https://threedesign.store/

I’m always interested in feedback though (especially if it’s a horrible idea).

[malcolmputer]

@CalvinStence

just a bit

How much is the BOM cost on it? If it’s just a little over we could probably take it to a vote.

What do you guys think of this Open Source CNC project:

Specs - Ghost Gunner

They sell them assembled for $2100, so I’d assume the BOM cost isn’t too bad.

I like the idea of an enclosure, I’d rather have it be contained than be an open frame style.

[tmc4242]

Also if I were to pursue such a machine, 8" of vertical travel would be a minimum. Maybe 12".
( I’ll have to do some digging to find the actual minimum. )

[malcolmputer]

@Kentamanos

(especially if it’s a horrible idea).

Not enough Z travel. 2.5" makes it a router not a 3 Axis IMO.

[tmc4242]

I see we’re on the same page…

[CalvinStence]

Joking aside though, this is something I have done quite a lot of thinking about and even more research on. I think we will be guided by what we really desire. Is it a milling machine we want, or is more of a router OK?

In my experience these are the challenges of designing a milling machine that can precisely machine whatever you want, albeit slowly:

1. The spindle–how to get one at all, how to hold tools without manually resetting each time which is such a pain in the ass
2. The spindle drive–How to have both speed, power, and torque required for the application
3. The machine base–how to get a precise, stiff, and heavy object to mount the motion assemblies to and dampen vibration
4. Overall stiffness, especially in torsion–Very general number here, You really want to be in the 20+ N/micrometer range to do not go crazy. 100+ N/micrometer is where things start to get good.
5. How to get big frame pieces that are precise and not infinity $ - We can’t really do cast iron lol
6. Precision - A mill can be fun without it but it’s much more useful if it can be geometrically aligned and be repeatable
7. Axis drives - What motors to use to get torque and speed
8. Power - how to get it to the motion components/spindle
9. Control - how to give commands to the motors and spindle

In my travels I have found a few methods to address these issues on a machine of the approximate scale we are discussing, but with capabilities closer to a milling machine than a router.

1. 5. Cheap R8 spindle head from a Sieg X2 with TTS holders (i have one I can donate, they are something like $100-150)
2. Cheap Chinese servomotors OR treadmill motors (varies but maybe $200 is expected)
3. Granite surface plates with threaded inserts epoxied in for a base OR epoxy granite (you can get tiny ones for about $50)
4. Double column fixed gantry construction: Massive gain in machine stiffness with minimal use of material due to effective double fixed end columns caused by gantry acting as a virtual fixed surface (cost comes in the form of item 5)
5. Laser cut steel and Diamant metallplastic (micro steel bearing ball impregnated epoxy which sets to 30% of elastic modulus of steel) bonded and bolted assemblies for stiffness, full contact on non ground parts, can be incredibly cheap. Just need to keep them from rusting. (maybe $300 if we fully machine all the parts and don’t laser cut them)
6. This is really difficult to do cheaply. Maybe we could make ways from surface ground cast iron bars and turcite. That would be a really fun project. But the easy way out is cheap Aliexpress linear guides and ball screws. This would run us about $400.
7. Ideally closed loop steppers for a mill are desired but they get expensive.
8. DC power supplies are usually needed here and depending on the power output they can be cheap or not, I estimate this will be about $150
9. Crap PC with parallel port can work, Ideally you want a motion control board like a centroid acorn, but you can just send steps over a parallel port. Maybe $100 max here

In total this is about $1200 give or take. This machine would not exactly be in the spirit of what you suggested Malcolm (I suspect it could take far larger depts of cut) but I believe it is minimum viable for a milling machine and could be surprisingly capable. If we can cut any costs it will be in trading off the shelf parts for self made things, which could make it more fun. Or we could move for more of a router type machine, which can be very capable even in steel if it is well designed.

Curious to see others takes on this.

[CalvinStence]

Not exactly sure but it’s at least $9k, probably more. That thing is 20in X by 15 in Y and 13 in Z (travels that is), 2000 pounds, the 5hp spindle motor and drive with shipping is over 1k LOL

[CalvinStence]

Oh I wish!

[tmc4242]

@CalvinStence

Crap PC with parallel port can work, Ideally you want a motion control board like a centroid acorn, but you can just send steps over a parallel port.

Just NO.

Modern control board with USB or Ethernet.

Parallel port is DEAD.

[malcolmputer]

@CalvinStence

Is it a milling machine we want, or is more of a router OK?

For me it’s mill or nothing. I’m not interested in doing a router. The specific project I have in mind that I’d like to be able to mill is about 4 inches tall and roughly cubic, and I don’t really do anything that is more of the multicam style really large but not tall project.

@CalvinStence

Not exactly sure but it’s at least $9k,

Oof. I’d rather deal with some loss of repeatability and stability.

@CalvinStence

2000 pounds

That’s a big issue as well. I’m looking for the $300 3D printer solution to the $200k SLS industrial printer problem here. I know I’ll be making some trade offs, but Wife Acceptance Factor is a consideration. I can’t take up a portion of my garage and run three phase out there, and if I could, I’d get a Bridgeport and skip the CNC side of things.

[tmc4242]

@malcolmputer

For me it’s mill or nothing. I’m not interested in doing a router.

Same here.

[michaelb]

I like the idea. We might be able to do some extra things in a group that we couldn’t alone, like having a shop water jet/laser cut some parts and only doing the finish machining. 6" cubed isn’t large, but I think it’s a good size for a first crack at a home built mill.

I fully agree that I’d rather have a mill than a router. Specifically with a relatively low speed spin (Small machine means small tools, but I don’t need a screaming 24k router spindle).

I don’t really require a tool changer or even quick change tooling. If the spindle has an ER collet nose and I have to manually adjust and live within the ER system, I think I’m ok with that. Drawbars and tool changers seem to get complex, expensive, and take up space quickly.

Always loved the idea of a machine based on a granite plate. I’m not sure if it’s within the budget but it’d be really cool.

I’m a fan of linear guides and ball screws. I’m not sure I want to fiddle with tiny little ways and gibs.

[TBJK]

If, & this is a big if, if you wanted mass in your favor, in theory, you could use a casting from an existing machine to mount off of. You are looking at scrap wise maybe 100$ a ton. This would give you a really heavy mass to mount things up to.

[malcolmputer]

@TBJK

This would give you a really heavy mass to mount things up to.

That could be neat! I don’t want mine to be too heavy to be lifted by one or two people, but once we’ve got them tuned in as a group, having it permanently installed somewhere weight won’t matter too much.

[CalvinStence]

I am definitely not proposing the first design I posted; that was just a joke! It is the umpteenth design iteration that began as the same idea this thread is looking to do. I had started thinking at the scale of a 12x18x3 inch surface plate with the Sieg X2 spindle–quite small and easy to handle, but still quite stiff. But I eventually settled on 2 24x36 plates glued together… just a bit bigger!

For sure though, a milling machine is certainly more desirable. Routers are just not very useful unless you’re trying to make huge aluminum plates.

But it is definitely not required to have a milling machine and have it be 9k. Again I think $1200-1600 for something barebones but functional is about what we can shoot for if we get creative. The methods of the bigger design I am working on myself could still apply when making a much smaller machine that is still capable of doing reasonable milling tasks (ridgid tapping, quasi-efficiently milling steel-etc). We can make it cheap because we can use methods, such as plate and adhesive construction, that aren’t practical for production of machine tools for sale.

Hearing what we want to do, if a 4" cube is the target workpiece size, here’s a curve ball for us depending on the features the part would need:

4" cube requires probably more than a 6" travel in Z. It needs to be held somehow and you need to have clearance for tools in the spindle, especially if you plan to drill large holes in it. For example, here my design has a 16" spindle nose to table distance:

But with a somewhat short tool and small vise, I only have 9" clearance to the top of the vise:

This gets much less if I have a drill chuck and a 1/2 drill:

There are many creative ways around this problem, especially if you are willing to sacrifice some stiffness and clearance of non Z-axis spindle components and the workpiece. But it’s something to keep in mind. X and Y are less of a challenge but extra Z clearance and travel is worth its weight in gold.