Looking for Advice: Desktop CNC Mill for Sale

[finaltheorem47]

Hello, I am in the process of developing a large desktop CNC mill for prototype work with some friends with the hopes to sell and I would like some feedback on the design. We’re hoping to target the makerspace, hobbyist, and prototype shops.

The CNC mill features a gantry style structure that is made from fully cast gray iron with all axis’s using linear guides and ball screws with hybrid stepper motors (steppers with encoders) with a target weight of approximately 350 lbs. It will use a bt30 spindle with power draw bar, and be offered with a 110V AC spindle motor capable of 6k rpm and an upgraded bt30 spindle, 220V AC spindle motor capable of about 12k rpm that will not trip and breakers. We are considering dropping the 110V option. We are attempting to minimize the machine’s weight by using FEA to selectively remove material while maximizing rigidity. The machine will be just narrow enough to fit through a standard single width door.

For electronics, we are planning to use a FPGA motion controller capable of S-curve acceleration (higher order motion planning). User interface will be through a 24” touchscreen LCD monitor. We are also planning on creating an optional enclosure to sell with it also. Tool height setting will be made easier with an integrated tool height setter.

We believe that quality control is essential and plan to assembly the entire unit in the United States. Spindles will be tested on a dyno and motion systems will be tested with a Renishaw ball bar to ensure performance. What do you think? We are also considering selling the machine just as a kit only, basically all the parts but unassembled. What would you be willing to pay for a machine such as this if you were considering purchasing one?

We’re also considering a lower cost option using an ER20 spindle but would not have a power draw bar and would require you to change tools manually and set the tool height each time.



Here is a picture of an earlier iteration for reference. The newer iteration has a different gantry/ z-axis profile that is still being developed.




Similar Threads:

• Desktop Mill advice needed
• Build Thread- Small desktop mill - New builder, advice appreciated
• Need Advice on a Desktop Mill

[awerby]

What is this machine intended to cut - steel or softer materials? Why the odd tilt in the Y axis beam? That seems to make building it more complicated without any obvious performance gain. And those leveling feet would work better if there was a way to secure them to the floor or desk. How much are you going to be able to build it for?

[finaltheorem47]

Hello Andrew, we intend for the machine to be able to cut all materials including steels. Because there is no flood coolant and chips are evacuated by compressed air, materials such as wood can also be machined relatively easily without getting dust stuck and glued to every surface.

The spindle (not the one pictured) is a separated robust bt30 spindle/ac motor system connected via belt drive. The spindle is one of the most expensive components on our whole machine as when we say we plan for it to cut steel, we really mean that we plan for it to cut steel. The cheaper 1.5/2.2/3 kW integrated spindles you can purchase from china will not last long at all cutting steel and harder materials. Not only does machining drastically steel reduce the lifespan of those spindles but they also don't have enough torque to hog through steel at the lower rpm range (700-2500ish) that you'd want when cutting harder materials. At best you can expect to take light passes with it. The spindle/motor we selected will allow for slotting operations with a 1/2" endmill no problem.

The Y axis beam tilt is gone in the latest iteration, I'm sorry I can't post it yet, I'll put it up asap. It was originally added to attempt to reduce torsional and bending strain the gantry would experience under load.

Regarding the price; we are still doing market research and trying to get a feel for what people believe is most important in a CNC that they chose to purchase and use. Because of that, components/sizing/weight/capabilities/features are all subject to change and we're not confident on stating a price. We're hoping to hear "We want this, this and this. This feature isn't important but this one is really important. If you could build a machine to do this I could see myself paying $XXX for it."

Fantastic feedback, thank you for your comment. It's great to get an outside opinion! Let me know if you have any more questions.

-Joe

[atomarc]

Joe,

My two cents..using compressed air for chip evacuation and to a degree, cooling is not a grand idea. I think you are going to want coolant and a enclosure to make the machine viable for anything other than cutting balsa wood.

Stuart

[awerby]

I agree with Stuart; blowing fine metal shavings around the room isn't a particularly good idea. It's not especially helpful for the metal-cutting process, and those needle-like chips can be dangerous. Flood cooling is good, but it makes a mess, especially if you're blasting the coolant hard enough to dislodge the chips. That's why VMCs typically are built with a water-tight enclosure. If you don't want to go the VMC route, then maybe mist coolant would be an acceptable compromise, although you'd have to brush the chips away by hand. I don't think a combination wood and metal machine really makes a lot of sense, although wood routers and metal-cutting mills do superficially resemble one another. I'm not sure what spindle would really work well at 700 rpm cutting steel and also cut wood at 20k rpm or so. The feedrates are also a lot different; wood wants to be cut fast, steel's cut slow. Maybe you could offer the same basic frame and fit it out one way or the other, offering a dust shoe for wood or a flood coolant system, a fast spindle for wood or a slow one for steel.

Any plan for a ATC? That's the feature everybody's going to want, but I don't see it in the drawing, or any space to put the tool rack. .

[dharmic]

An air clear/cool even with a low rate of mist for a little lubrication works really well BUT I wouldn't be using it if I didn't have an enclosure on my machine because it does make an awful mess.

That said, my machine (that cuts mostly aluminium and HDPE with some occasional work in steel and timber) is a 'desktop' 6040 style and there's no way I'd have it in the office. Even with the enclosure, it's just too noisy and messy.

[ger21]

What would you be willing to pay for a machine such as this if you were considering purchasing one?

That's impossible to say without knowing the final specs and cost, and seeing real world results from the machine.

For electronics, we are planning to use a FPGA motion controller capable of S-curve acceleration (higher order motion planning).

Are you creating your own control, or using a TinyG?

[finaltheorem47]

Here is a reply from one of my partners who is helping with the mechanical design and focusing on the electronics.

Hi everybody, thank you so much for the opinions and responses. As far as the cooling goes we have plans to include a mister cooling unit and the unit alone right will come with a chip pan and the enclosure will have 3 doors 1 front and 2 sides and they will be running on linear rails with catches to keep them close. The enclosure will be a fully sealed one I had experience with knees mills and cleaning up everything was not my favorite since chips got everywhere. As far as materials for the enclosure it will use 18 gauge sheet metal and the windows will have plexiglass on them. The lighting will be LED flood lighting.

As to the response to motion control; we are planning to team up with some people to develop our own. We are not using TinyG as I found some random issues of just the USB breaking our controller so we will use an Ethernet port to send and receive data. We also want to plan for future expansions and capability and the tinyG is a limiting factor. The entire machine will be ran from a RPi or similar SOC. The motion controller we will be using has a STM32 bit microcontroller that is coupled with a FPGA that does the actual step generation, so we can get cleaner generation without over loading the microcontroller. I had experience running 3d printers and it seem a lot of 32-bit system are limited to a 120 to 150 khz range and I would like to avoid that bottle neck from the get go.

As far as the ATC is considered we will be having a small rack style (3 to 8 tools) one and as we develop as a company we plan to possibly launch a side mount that would go with this version and its bigger brother if everything worked out.

Stuart: We are designing an enclosure to go with it but were thinking about making it optional for those who wanted the purchase the machine at a discount. We are not however planning currently for flood coolant to be used. We believe nothing/compressed air/mister satisfies our objective of a prototype level cnc mill w/out needing to push it towards having flood capability. Flood coolant needs to be maintained (bacteria and stuff can grow in it even) and from personal experience I don't like to let it sit idle for extended periods (a hobbyist for instance).

Andrew: Great idea on offering two versions depending on the materials you want it to cut and I think we might be able to do that if people are interested. My partner answered your question regarding the ATC above.

Dharmic: Agreed, I personally love mister; it works fantastic. Milling metal in my experience makes a mess even when you don't use compressed air haha especially with a facemill. We were thinking about what we could do to make it low noise if we really tried. We were wondering how quiet we could make the machine if we stick it in an enclosure that has been insulated for sound. We figure flood coolant would very quickly destroy any insulation we install, but since we're trying to avoid flood coolant in the first place, it may be feasible. Regardless, It's something we definitely plan to try out.

Ger21: True, I agree completely. It's like saying "I want to sell you a car sight unseen. What will ya pay?" I believe asking that question was a poor idea; I think asking "what would you want" is better, and we will try to figure out a price based on your specifications. The tinyG question is answered in the above quote.

[cyclestart]

From the base up the design brings this to mind.

and the build thread on these forums
https://www.cnczone.com/forums/bench...27-cnc-24.html
I like the idea bridge mills over C frames but suspect they cost a bit more to build.

[SCzEngrgGroup]

From the base up the design brings this to mind.

and the build thread on these forums
https://www.cnczone.com/forums/bench...27-cnc-24.html
I like the idea bridge mills over C frames but suspect they cost a bit more to build.

And that build has been in-process for well over 10 years now....

Regards,
Ray L.

[awerby]

If you're building a complete enclosure, then I'm not sure why you're not offering a flood coolant option with it. Re-cutting chips is one sure way to dull cutters and degrade cut quality. Yes, you can open the door, inhale some mist, get in there with your chip brush and sweep them away every few minutes, but that's not a lot of fun, and doesn't work as well as simply blasting them out of the cut with the coolant. And most VFDs have a door interlock for safety's sake so that doing that would stop the program and possibly leave a dwell mark (or worse).

[cyclestart]

And that build has been in-process for well over 10 years now....

Yeah and that build thread also goes off in a few tangents involving the builders a other projects. It was the mention of castings in post #1 that made me think of it.

Joe, I've seen machines similar to yours on Alibaba or Aliexpress but can't remember seeing any American sellers. Certainly can't think of any American manufacturers

EDIT/ A company in California has a 250 lb desktop cnc bridge mill. On their youtube channel they claim "ActMachines manufactures the best desktop CNC machines in United States". It may be built in America or the manufacturer may be American based depending how you parse that statement.

[finaltheorem47]

From the base up the design brings this to mind.

and the build thread on these forums
https://www.cnczone.com/forums/bench...27-cnc-24.html
I like the idea bridge mills over C frames but suspect they cost a bit more to build.

Fascinating build. Thank you for sharing, I have not seen this before but it looks very similar. Has the creator (W. Smith) had any issues of excessive strain on x vector loading?

If you're building a complete enclosure, then I'm not sure why you're not offering a flood coolant option with it. Re-cutting chips is one sure way to dull cutters and degrade cut quality. Yes, you can open the door, inhale some mist, get in there with your chip brush and sweep them away every few minutes, but that's not a lot of fun, and doesn't work as well as simply blasting them out of the cut with the coolant. And most VFDs have a door interlock for safety's sake so that doing that would stop the program and possibly leave a dwell mark (or worse).

We considered flood cooling but leaned away from it because of how messy it'd be. We were considering trying to make our niche as a low noise machine, using insulation in the enclosure to limit sound and that would not be easy to implement with flood coolant. We also assumed that our target market of hobbyist/maker space/research company would be less inclined to use flood coolant when prototyping, but if there is support for it we will reconsider our decisions.

Yeah and that build thread also goes off in a few tangents involving the builders a other projects. It was the mention of castings in post #1 that made me think of it.

Joe, I've seen machines similar to yours on Alibaba or Aliexpress but can't remember seeing any American sellers. Certainly can't think of any American manufacturers

Yes, we've seen many of these machines for sale but we believe we will have an edge over them due to the time we're spending in the design and testing we plan to put the machine through before bringing it to market.

We plan to machine everything in house to maintain QC standards.

[cyclestart]

Has the creator (W. Smith) had any issues of excessive strain on x vector loading?

You could post to his build thread or even send a PM, he would probably be happy to discuss his projects.
Good luck with your product.

[finaltheorem47]

Hey guys, it's been a few months and we're just days away from a working functional prototype!

We had some trouble with our original spindle which was bad when we received it wasting time. We tried to fix it but were not satisfied which also wasted time. We then built our own which took time and ended up constantly delayed between steps in the process so we eventually ordered a nice off the shelf BT30 to use which will be coming in soon. We were very excited to see the results of how rigid our fully cast iron frame was. There have been surprisingly few roadblocks other than the issues regarding the spindle.

Here is a teaser picture of what we believe showcases medium machine level capability cut in steel with a 3/8" endmill, full slotting WOC, .25" DOC. Our goal is to be able to double the DOC to about .5" as far as the maximum ability of the machine's motor goes.




If you are interested in our project and want to keep a tab on us for information including our plans for launch, please either email us at:

Team@vulcanmachineco.com or Vulcan Machine Co.

We plan to send out an email every week or two with our progress to whomever is interested.