[DuxEtCapital]

Hi everyone!

This is my first post, but I've been lurking the forum every now and then for a while.

I am playing around with the thought of building a VMC from scratch. I have all the tools and machinery available; CNC-mills and lathes, welding tables, materials etc. I'm employed as a CNC-operator (apprentice at the moment).

Before anyone ask why I would like to build my own if I already have everything available, the answer is simply I would like to build one to fit besides my computer desk at home. I'm not aiming for a small 200x300x100mm CNC mill - if I'm doing this I'm going all in. I am not looking to carve things in wood or plastics. I want a VMC being strong enough to cut stainless and other hard steels. I'm probably going to work a lot with aluminium - but I know I will regret not making it strong enough for harder metals.

Before anyone points it out, I would like to say that I have limited knowledge about the need for torque performing these tasks - which is why I am opening a thread. My hope is that some of you wise souls in here can shed some light on some aspects for me. Pardon me if I seem a bit to ambitious, and if it seems like I don't know what I'm doing. I'm still learning CNC, and I'm a talented CAD-designer. Everything is still conceptual at the moment, so I don't have any CAD drawings yet. Right now I'm just collecting information!

I'm picturing a budget of around 10 - 15 000 USD, with investments made over time. I am not rich, but I make an okay living.

I'm thinking I would weld the frame in 100x100x5mm squared steel tubing, with crossovers for stability. To increase the work-area, I want to get the spindle doing the moving, while the table stands still. I know most CNC-mills have the spindle moving only in Z-axis to increase spindle stability - but I have some ideas.

I am thinking of using 4 HGW65 linear rails for each axis to increase stability - combined with some heavy duty ball-screws with double nuts to keep backlash to an absolute minimum. If I make it so that I can adjust the pressure on the rails, I can adjust it until it's dead-tight.

Regarding the axis control I started thinking about steppers, but after what I've understood these are inferior to servos. I've been looking around at some servos at eBay that I suspect should do the trick. With the little knowledge I have I see that it's capable of 15NM, which seems stronger than most servos for CNC online. I've understood that most DIY'ers build routers for wood and plastics, so I guess I need something stronger than usual.

Would this do the trick for something like this?
https://www.ebay.com/itm/3-8KW-NEMA5...EAAOSwWORbmI2s

Eventually, would the use of matched speed reducers be sufficient to gain more power? What ratio would you suggest=

Regarding the spindle, I've contacted a chinese spindle manufacturer, and I think they have what I'm looking for. They can also give me CAD-drawings of the spindle so that I can design the VMC to fit it before I order. I have the following demands for the spindle; water cooled, strong enough to cut steel, high quality bearings (preferably ceramic), automatic tool change, and ready to accept BT-30 tool holders. They've sent me a suggestion to a 7.5 Kw spindle that they say should do the trick. Picture is attached! Thoughts on this?

By all means guys; if you think I'm a way to ambitious dreamer with his head up his a** - I expect to hear it on a public forum as large as this. I've been using forums for too long to expect anything else. But please, any tips will be taken to consideration with gratitude.

Thanks for reading!

[awerby]

Hi everyone!

This is my first post, but I've been lurking the forum every now and then for a while.

I am playing around with the thought of building a VMC from scratch. I have all the tools and machinery available; CNC-mills and lathes, welding tables, materials etc. I'm employed as a CNC-operator (apprentice at the moment).

[Have you got metal-casting equipment set up? All the VMCs I know of are based on heavy iron castings, not weldments.]

Before anyone ask why I would like to build my own if I already have everything available, the answer is simply I would like to build one to fit besides my computer desk at home. I'm not aiming for a small 200x300x100mm CNC mill - if I'm doing this I'm going all in. I am not looking to carve things in wood or plastics. I want a VMC being strong enough to cut stainless and other hard steels. I'm probably going to work a lot with aluminium - but I know I will regret not making it strong enough for harder metals.

Before anyone points it out, I would like to say that I have limited knowledge about the need for torque performing these tasks - which is why I am opening a thread. My hope is that some of you wise souls in here can shed some light on some aspects for me. Pardon me if I seem a bit to ambitious, and if it seems like I don't know what I'm doing. I'm still learning CNC, and I'm a talented CAD-designer. Everything is still conceptual at the moment, so I don't have any CAD drawings yet. Right now I'm just collecting information!

[Torque is important, particularly low-speed torque for the spindle, but the main issue is rigidity. Most of us here are building routers of one sort or another, which at the top end can handle aluminum. But something that can cut stainless and other hard metals effectively is another thing entirely,]

I'm picturing a budget of around 10 - 15 000 USD, with investments made over time. I am not rich, but I make an okay living.

I'm thinking I would weld the frame in 100x100x5mm squared steel tubing, with crossovers for stability.

To increase the work-area, I want to get the spindle doing the moving, while the table stands still. I know most CNC-mills have the spindle moving only in Z-axis to increase spindle stability - but I have some ideas.

[That's less rigid than a fixed-spindle moving-table design. Mills are all done the other way for a reason.]

I am thinking of using 4 HGW65 linear rails for each axis to increase stability - combined with some heavy duty ball-screws with double nuts to keep backlash to an absolute minimum. If I make it so that I can adjust the pressure on the rails, I can adjust it until it's dead-tight.

[I'm afraid you'll find that keeping all those rails absolutely aligned with one another is a total nightmare. If they're just a little off in any direction the whole thing will bind up. There's no way those welded tubes are going to be flat, rigid or stable enough for all those rails to mount orthogonally. And doubling up the nuts on a ball screw and loading them together makes it harder for the motors to turn them. "Dead tight" might as well just be dead. Less resistance to motion is the main reason you use ball screws in the first place. It's better to use a single nut and oversize balls.]

Regarding the axis control I started thinking about steppers, but after what I've understood these are inferior to servos. I've been looking around at some servos at eBay that I suspect should do the trick. With the little knowledge I have I see that it's capable of 15NM, which seems stronger than most servos for CNC online. I've understood that most DIY'ers build routers for wood and plastics, so I guess I need something stronger than usual.

Would this do the trick for something like this?
https://www.ebay.com/itm/3-8KW-NEMA5...EAAOSwWORbmI2s

[Servos would probably be better than steppers for a machine like you describe, although I don't know anything about those particular ones. You're in Australia?]

Eventually, would the use of matched speed reducers be sufficient to gain more power? What ratio would you suggest=

[Servos run a lot faster than steppers, so speed reducers are a good idea to get into their sweet spot. But you'd need to find ones that don't introduce backlash.]

Regarding the spindle, I've contacted a chinese spindle manufacturer, and I think they have what I'm looking for. They can also give me CAD-drawings of the spindle so that I can design the VMC to fit it before I order. I have the following demands for the spindle; water cooled, strong enough to cut steel, high quality bearings (preferably ceramic), automatic tool change, and ready to accept BT-30 tool holders. They've sent me a suggestion to a 7.5 Kw spindle that they say should do the trick. Picture is attached! Thoughts on this?

[I doubt a 36k rpm spindle will have much torque at all at the speeds you'd use for cutting steel, unless you're using very small cutters. It would be OK for wood, though.]

By all means guys; if you think I'm a way to ambitious dreamer with his head up his a** - I expect to hear it on a public forum as large as this. I've been using forums for too long to expect anything else. But please, any tips will be taken to consideration with gratitude.

Thanks for reading!

[It's okay to dream, but try to wake up and pull it out of there before you start spending serious cash. If what you really want is a good working machine as cheap as possible, you'd be better off looking for a good used VMC with a blown or obsolete controller and retrofitting it with something more state-of-the-art. At least in the US, one can often find them pretty cheap from people who basically just want their floor space back for a newer or better one. But if you're mainly interested in proving all the other mill designers in the world wrong by showing how it can be done so much cheaper and easier, go for it, and post a video when you're done of this thing chewing through stainless steel like a hot knife through butter, leaving a surface that looks polished already...]

[dharmic]

I think most of us have been here and shared this dream. A few have gone on to spend bulk dollars for the great feeling of a DIY well done leaving them with less machine for more bucks than a 2nd hand rebuild.

I want to home in on one thing, though, and that's the ergonomics of a machine inside the domicile: you lost me at "to sit beside my computer at home". Presumably in a study or your bedroom? I have all my machines in the shed except for the 3D printer which does live in the same room as my computer in the house and even that is too noisy most of the time to have running and still be able to concentrate on a job at the computer. Even with a full enclosure, the printer generates a lot of mess on the floor and leaves dodgy fumes in the air. A mill cutting stainless? Forget it, the noise will kill you and if anyone else lives in the house they'll kill you too. That's just the machine, then add the noise of the shop air you invariably need to run them effectively (if not for the collet holder then at least for clearing chips etc).

Then move onto the mess these things make, even with an enclosure you'll have chips and coolant all over the place. Your floor, wall, bench and probably computer too will be stained beyond redemption in short order.

Then consider the stank from coolant, lubes, hot metal and so on. It's fine, it's better than fine, it's the smell of heaven in the context of a workshop. But in a small room in a house or apartment where you can't leave it behind after a job? You'll learn to hate it in short order.

There's a spot in my shed where I could put the printer, and I don't, because I can put up with the minor noise and mess issues for the sake of having it close to keep an eye on when it's running. Thinking outside the box, maybe it'd work for you too: forget the CNC and instead have a printer at home for prototyping, then if you get to a form you like and want to do in stainless, chuck the post processor output on a thumbdrive and sneak it through at work when a machine's idle...

Don't despair - many of us posted something similar to this thread and we all got slammed and, a few months later, most of us were grateful for being shut down early

[footpetaljones]

1. How much power do you have to drive this machine? You're headed in the direction of needing a dedicated 3 phase circuit.

2. The torque you need depends on how much weight you want to move and how fast you want it to get there. I know many industrial machines might use a 3 kw motor to move an axis at 1500 IPM, but that big of a motor is only needed when that axis has more stacked on top of it. Direct drive, no reductions necessary. Those same machines generally use 32 and 25 mm ballscrews.

3. Do not use 4 rails. If that much rigidity is needed, move up to 2 rails with rollers.

4. For mills that don't use a typical XY on table/ Z on column design, look at DMG Mori's designs. A lot of their youtube videos show exploded views of the machine. Brother's pallet changing D&T machine has a static table with XYZ all on the column.

5. For steel you need torque, not RPM. 12K RPM is a good middle ground for some steel and aluminum jobs. It will be difficult to find a motorized spindle with RPM at 12K.

6. In my experience, Chinese companies will tell you anything you want to hear.

7. I agree with dharmic. Either it can fit besides your computer, or it can be a VMC. Clarify your goals, but keep in mind a Tormach 440 can cut stainless. Slowly and not terribly accurately but it will cut it.

8. 10-15000 USD is definitely doable for a 3 axis mill the size of a RoboDrill or Brother. $1500 for steel tubing/bars, $2000 for DMM servos, $2000 for Chinese rails/ballscrews, $2000 for a Chinese servo spindle and drive, $1000 for a Chinese BT30 spindle, $1000 for metrology equipment to read tenths, $500 for a mini PC for LinuxCNC, $1000 for electrical components, $500 for pneumatic components. As Andrew said, though, it's generally cheaper and easier to buy a used machine.

[mick41zxr]

Half way through building pretty much as you describe, however I would suggest as others have suggested & retrofit an existing machine. they seem to be plentiful for reasonable coin states side.
Otherwise look for an RF45 clone with a BT30 spindle. Built with closed loop steppers would be a god compromise.
Careful as costs can get away from you. If you can get away with doing a few foreign orders on your work machines may save some.
Regards,
MB

[hanermo]

Go ahead !

It will be an interesting and educational and expensive journey.
There are 3-4 relevant samples of people milling steel with shop built tools. Afaik.
The wall thicknesses are typically 1/2" or more.

I am re-finishing (version 5) my big VMC.
Parts cost alone == 20.000€ in bulk import wholesale costs.
Total current value of tools+parts == 100.000 €.
+ 17.000+ hours of work.

Certainly a small VMC can be made for 15.000$ in parts..
.. if you aim for cheap vs good.

Good results from rigidity.
Thicker screws, bigger linears, thicker section size flats.
Bigger section sizes.

All modern machine tools embody about 2% load vs capacity.
So do all good lathes, since forever.
My VMC with == 1000 kgf force on x,y has 50.000 kgf linears on the z axis.
35 mm hiwin hgw, and doubled rails on the front. 32 mm screws.

[jgwentworth96]

As mentioned above your biggest barrier to getting this in a home is power. If you're looking at having something with enough torque to cut hard materials you will easily exceed the amount of power you can reasonably run off of 220V single phase. Unless you have 3 phase power already run to your home it's going to cost thousand to tens of thousands to get it wired to your house (if its available at all). The 7.5 kW spindle alone would pull 35 amps on 220V 1ph.

[Andreiir]

First of all , u need to see what precision u need , if u need high precision , u need go epoxy granite , if u dont need that much precision u can go with steel frame.
Noise is not a problem with a very rigid vmc , it wont sound at all , and if u build it with eg , it will make low noise .
Beside this u can make an enclosure isolated , and it will be ok , for all sounds that will bother u .
Don"t listen to people that want get u out of line with ur dream , just do it and leave others with they dreams .
With that money u can buy an vmc , and retrofit , but is not like u do it , u learn a loot of thinks , if u make the vmc . And u will know all the tricks for building one vmc , but u need tools for this , and help of other machinist with a good vmc , for aligning the rails and ballscrew .

[DuxEtCapital]

I think most of us have been here and shared this dream. A few have gone on to spend bulk dollars for the great feeling of a DIY well done leaving them with less machine for more bucks than a 2nd hand rebuild.

I want to home in on one thing, though, and that's the ergonomics of a machine inside the domicile: you lost me at "to sit beside my computer at home". Presumably in a study or your bedroom? I have all my machines in the shed except for the 3D printer which does live in the same room as my computer in the house and even that is too noisy most of the time to have running and still be able to concentrate on a job at the computer. Even with a full enclosure, the printer generates a lot of mess on the floor and leaves dodgy fumes in the air. A mill cutting stainless? Forget it, the noise will kill you and if anyone else lives in the house they'll kill you too. That's just the machine, then add the noise of the shop air you invariably need to run them effectively (if not for the collet holder then at least for clearing chips etc).

Then move onto the mess these things make, even with an enclosure you'll have chips and coolant all over the place. Your floor, wall, bench and probably computer too will be stained beyond redemption in short order.

Then consider the stank from coolant, lubes, hot metal and so on. It's fine, it's better than fine, it's the smell of heaven in the context of a workshop. But in a small room in a house or apartment where you can't leave it behind after a job? You'll learn to hate it in short order.

There's a spot in my shed where I could put the printer, and I don't, because I can put up with the minor noise and mess issues for the sake of having it close to keep an eye on when it's running. Thinking outside the box, maybe it'd work for you too: forget the CNC and instead have a printer at home for prototyping, then if you get to a form you like and want to do in stainless, chuck the post processor output on a thumbdrive and sneak it through at work when a machine's idle...

Don't despair - many of us posted something similar to this thread and we all got slammed and, a few months later, most of us were grateful for being shut down early

Would the smell, chips and fumes be a problem if the VMC has a full enclosure, with constant ventilation blowing everything outside? The chips would hit the walls, and fall down into the shute - and a conveyor belt taking the chips out into a 40L bucket? I mean even at work the chips is rarely a problem unless we use the blow-gun and chips go flying everywhere. I was thinking I could get around this using a fan-tool in the spindle. So that I can make a program to clean the table/workpiece after machining is complete.

I already have a printer at home which have served me for some time now. I spend enough time at work as it is, and would like to enjoy milling in the comfort of my own home. I totally get what you mean regarding sound though. I though I should completely isolate the enclosure with several layers, as well as isolating the office where it would be placed. This combined with as rigid frame as possible, as well as rubber feet - I hope to minimize it to a level where I will be able to live with it.

I don't have a garage - or a shed for that matter. :P

[DuxEtCapital]

1. How much power do you have to drive this machine? You're headed in the direction of needing a dedicated 3 phase circuit.

2. The torque you need depends on how much weight you want to move and how fast you want it to get there. I know many industrial machines might use a 3 kw motor to move an axis at 1500 IPM, but that big of a motor is only needed when that axis has more stacked on top of it. Direct drive, no reductions necessary. Those same machines generally use 32 and 25 mm ballscrews.

3. Do not use 4 rails. If that much rigidity is needed, move up to 2 rails with rollers.

4. For mills that don't use a typical XY on table/ Z on column design, look at DMG Mori's designs. A lot of their youtube videos show exploded views of the machine. Brother's pallet changing D&T machine has a static table with XYZ all on the column.

5. For steel you need torque, not RPM. 12K RPM is a good middle ground for some steel and aluminum jobs. It will be difficult to find a motorized spindle with RPM at 12K.

6. In my experience, Chinese companies will tell you anything you want to hear.

7. I agree with dharmic. Either it can fit besides your computer, or it can be a VMC. Clarify your goals, but keep in mind a Tormach 440 can cut stainless. Slowly and not terribly accurately but it will cut it.

8. 10-15000 USD is definitely doable for a 3 axis mill the size of a RoboDrill or Brother. $1500 for steel tubing/bars, $2000 for DMM servos, $2000 for Chinese rails/ballscrews, $2000 for a Chinese servo spindle and drive, $1000 for a Chinese BT30 spindle, $1000 for metrology equipment to read tenths, $500 for a mini PC for LinuxCNC, $1000 for electrical components, $500 for pneumatic components. As Andrew said, though, it's generally cheaper and easier to buy a used machine.

I have 220V. I was thinking I could use multiple power supplies - so that each servo, and the spindle, can get their own. Connected to different ports around the house. That way I can distribute the load, and get around the need for all power through one port. I don't have a 3 phase circuit at home, and getting one in my neighborhood would be next to impossible - or at least exceed the the cost of the VMC by itself. I have electrician friends that I hope could help me find a way to make the multiple power-supply idea work.

2. I'll have to put some more thought into how fast I would like the machine to move. At the moment I still don't understand how I should calculate this. I guess I was thinking I would just get more power than I initially need - just to make sure I don't lack any power. I've checked out some 32mm ball-screws with anti-backlash from Alibaba. Again - rather too strong, than too weak.

3. Got it. Are you thinking so I can adjust the rollers to adjust the pressure on the rails, thereby tightening it up?7

4. We have some DMG Mori's at work. I was intrigued by their design to start with, and thought I should do something up that ally to maximize work area.

5. I'm Norwegian, so English is not my mothers tongue. You say I need torque, and not RPM for steel - but 12K RPM is a good middle ground for aluminium and steel? How would I measure the torque in a spindle? By horsepower? Kw? Would eventually the spindle I found with 7.5 Kw be sufficient?

6. Indeed. This is my experience as well. However, this manufacturer has delivered spindles for quite some time now - and it seems their previous customers is satisfied with their products. I have been struggling to find other spindle manufacturers than Chinese that can deliver a spindle with the demands I have for it. I'm sure that there are other companies that produces this elsewhere - but it doesn't seem that they are priced, or advertised to sell to individuals or hobbyist. Therefore I'm thinking that the Chinese is my only alternative. If you have any suggestions, then please post!

7. If there was any other machine that fit my needs I could retrofit, then I would. However, my demands are specific, I have the funds to do it, and I would like to try.

8. Just to clarify; I would get the materials like steel and aluminium I need from my employer for next-to-nothing. So I have quite some savings there. As mentioned in OP, I'll make the investments over time. So if I stop at 15 000 or 25 000 USD isn't really that important.

Thanks for the input!

[DuxEtCapital]

Half way through building pretty much as you describe, however I would suggest as others have suggested & retrofit an existing machine. they seem to be plentiful for reasonable coin states side.
Otherwise look for an RF45 clone with a BT30 spindle. Built with closed loop steppers would be a god compromise.
Careful as costs can get away from you. If you can get away with doing a few foreign orders on your work machines may save some.
Regards,
MB

I need a spindle with a pull-bar to handle ATC. I have a concept on a tool-magazine that would be too much too explain. I live in Norway, and the variety of machines to retrofit is not very big. I have considered retrofitting - but there just isn't any alternatives to retrofit here. Therefore; I'll rather take my chances of doing it scratch-built.

Exciting build there, friend! Do you have a thread where you post updates? Would love to follow your work!

[DuxEtCapital]

Go ahead !

It will be an interesting and educational and expensive journey.
There are 3-4 relevant samples of people milling steel with shop built tools. Afaik.
The wall thicknesses are typically 1/2" or more.

I am re-finishing (version 5) my big VMC.
Parts cost alone == 20.000€ in bulk import wholesale costs.
Total current value of tools+parts == 100.000 €.
+ 17.000+ hours of work.

Certainly a small VMC can be made for 15.000$ in parts..
.. if you aim for cheap vs good.

Good results from rigidity.
Thicker screws, bigger linears, thicker section size flats.
Bigger section sizes.

All modern machine tools embody about 2% load vs capacity.
So do all good lathes, since forever.
My VMC with == 1000 kgf force on x,y has 50.000 kgf linears on the z axis.
35 mm hiwin hgw, and doubled rails on the front. 32 mm screws.

I'm pretty sure I'll be able to make mine for less than 100 000 Euro, and 17 000+ work hours. I'm not making a 5x6m VMC. The total size of it will probably be 1600x1200x2100mm (Length, Width, Height).

Could you post a link to put me in the right direction regarding the load vs capacity question? What defines your kgf (Kg Force?)? Can I convert the oz.lb from the servos though an equation of some sorts to calculate the Kgf with the different speed reducers? Again, pardon my ignorance. I'm trying to learn this.

[mick41zxr]

Yes power will be a problem, especially on the spindle. Faced the same issue on current build. Most dedicated spindle motors are 3 phase.
I am testing one now that is single phase 2.2kW 8000rpm.
That being said I have 15amp circuit supplying spindle & 10 amp circuit supplying controls & servo's. Both 240v single phase.
What sort of travel distances are you looking for?
Regards,
Michael

[DuxEtCapital]

First of all , u need to see what precision u need , if u need high precision , u need go epoxy granite , if u dont need that much precision u can go with steel frame.
Noise is not a problem with a very rigid vmc , it wont sound at all , and if u build it with eg , it will make low noise .
Beside this u can make an enclosure isolated , and it will be ok , for all sounds that will bother u .
Don"t listen to people that want get u out of line with ur dream , just do it and leave others with they dreams .
With that money u can buy an vmc , and retrofit , but is not like u do it , u learn a loot of thinks , if u make the vmc . And u will know all the tricks for building one vmc , but u need tools for this , and help of other machinist with a good vmc , for aligning the rails and ballscrew .

I would like to get it as accurate as possible. I've seen some DIY'ers on YouTube that have used EG - quite frankly I thought they did this to save money, not to gain rigidity. Would I be able to gain the same result if I were to weld the tubes with all tube ends intersecting? So that I can fill the frame entirely with EG? Or do I have to make it so that the EG will be in direct contact with linear rails and such? Worst case I could make a mold and cast the whole frame, but I think it would be easier to weld it given the machinery and tools I have available.

Regarding isolating the enclosure; my thoughts exactly. I am thinking I would make it completely isolated, as well as further isolating the room the VMC would be in. If the enclosure is completely sealed, I could give it good ventilation, so that all smells and fumes would be sucked right outside when it's generated. As long as I can make the doors seal properly with a rubber gasket, I think it will be alright.

Thanks brother! I'm pretty sure that this project is doable. At work I have pretty much any kind of VMC available, and plenty of colleagues willing to help with more years of experience than I have totally in life. We have Matec, Quaser, DMG Mori's, Haas Mori Seiki - you name it. Posts like yours makes my faith stronger - so thank you for that good sir!

[rowbare]

I suggest browsing through the forum threads and see what people have build that matches up with what you want. Pay close attention to whether or not the build was ever completed...

While it is a moving table design, this is a particularly nice build: https://www.cnczone.com/forums/verti...262-forum.html

bob

[mick41zxr]

With access to equipment you have your project is certainly possible.
I used multiple tube (10mm wall) boxed & welded together. 25mm flat bar for rail mounts. Head was multiple laser cut 10mm plate. Whole lot internally braced & welded as well, has been post weld heat treated & will be filled with epoxy when done.
Using 30mm linear rail with 25mm C3 double nut ballscrew. Some of my biggest expense was machining that had to be sent out as my knee mill too small.
Regards,
Michael
Attachment 403018Attachment 403020Attachment 403022Attachment 403024Attachment 403026

[mick41zxr]

Also in regards to enclosure don't stress too much.
On previous build biggest noise was the blow off valve on my air compressor & if a tool chattered in a corner.
Regards,
Michael

[hanermo]

The max kgf force is defined by what your servos can deliver in Nm, and your ballscrews can impart.
A typical Haas Vf series machine or ST lathe is == 1500 - 1600 kgf max force.
The specs are public online.
Using a 32 mm ballscrew.

I use 32 mm ballscrews and 35 mm hiwin linear guides on my vmc (y is still 20 mm for now).
Lathe has 25 mm hgw wide high load high preload blocks.
The bigger haas (over == 1m table) (and everyone ) use 45 mm+ linears on the bigger machines.
But they are real 25 kW output power industrial tools with a C frame.

Basically a 32 mm screw equates to == 1500 kgf max force and 25 kW max spindle power.
I expect to get about 2-3 kW usable spindle power from an ISO30 spindle on mine (able to do 7-15 kW on the spindle, theory).

Std import screw mounts are crap.
They are undersized and don´t use real AC bearings.

For example the yoke on 12x my lathe, 32 mm screw on x, is 70 mm thick tool steel.
About 2+ x D on screw is right for a yoke.

For reference every single exceptional lathe in the world is about 2% load / strength.
Monarch 10EE, Schaublin 102, Hardinge, Sharp (monarch copy), Wheeler, etc.

Nm in torque x screw rise converts to push force.
The actual formula is on the web.
Essentially 3Nm stepper x 1:3 belt has == 600 kgf peak push force on a 5 mm screw.
In use, only 30% of that.

A 400W servo has about double-triple the real push force vs the stepper and 5x the accuracy for == 130$ more cost per axis.
My VMC has 400 W ac servos, and lathe has higher end 750W servos, 690€ each axis.

The 750 W servos == 10 Nm peak deliver == 700 kgf push force direct drive iirc.

My max speed and acceleration on the lathe are vastly higher than I want.
I want very high resolution, very high accuracy, good repeatability, and good - hopefully excellent - rigidity.
Using about 30% of max acceleration and speed.

- Load vs capacity.
A HAAS DT type machine has theoretical linear guide capacity in several tons, about 20.000 kgf on 35 mm linears iirc.
The rated table load capacity iirc is under 1000 kg.
About 20:1.

If someone loads the table at one end, failing the blocks, the MT builder must still replace and fix them.
Even though it is against specs, policy, usage guidelines.
Because the MT builder cannot afford a negative public meme on "weak" machines.

D:
I used to work as the Haas Spain commercial manager, under Hitec Mexico.
Nr 1 dealer of Haas machines in the world.
I have nothing except admiration and positive experiences from my time with them.
Bill and Andy are great guys, honest, fair, smart.

I looked at about 2000 machines in 200+ shops in 5 countries.
We in Hitec Spain had special training from the best guys in the world.
In 2013 the Haas Spain outfit was the nr 1 dealer in the world.

And in 2011/2012 I sold the best, most high-spec haas machines in europe, for the best margins, with very very satisfied customers.
65+ of them for about 5M++€.

Dux .. just using cheap 400W ac servos will be fine.
These will be more than you need for 5kW+ spindle power.

I'm pretty sure I'll be able to make mine for less than 100 000 Euro, and 17 000+ work hours. I'm not making a 5x6m VMC. The total size of it will probably be 1600x1200x2100mm (Length, Width, Height).

Could you post a link to put me in the right direction regarding the load vs capacity question? What defines your kgf (Kg Force?)? Can I convert the oz.lb from the servos though an equation of some sorts to calculate the Kgf with the different speed reducers? Again, pardon my ignorance. I'm trying to learn this.

[hanermo]

Excellent post.
Very much in line with my experiences and all those who mill steel.

With access to equipment you have your project is certainly possible.

I used multiple tube (10mm wall) boxed & welded together. 25mm flat bar for rail mounts. Head was multiple laser cut 10mm plate. Whole lot internally braced & welded as well, has been post weld heat treated & will be filled with epoxy when done.
Using 30mm linear rail with 25mm C3 double nut ballscrew. Some of my biggest expense was machining that had to be sent out as my knee mill too small.
Regards,
Michael

[dharmic]

Would the smell, chips and fumes be a problem if the VMC has a full enclosure, with constant ventilation blowing everything outside? The chips would hit the walls, and fall down into the shute - and a conveyor belt taking the chips out into a 40L bucket? I mean even at work the chips is rarely a problem unless we use the blow-gun and chips go flying everywhere. I was thinking I could get around this using a fan-tool in the spindle. So that I can make a program to clean the table/workpiece after machining is complete.

I already have a printer at home which have served me for some time now. I spend enough time at work as it is, and would like to enjoy milling in the comfort of my own home. I totally get what you mean regarding sound though. I though I should completely isolate the enclosure with several layers, as well as isolating the office where it would be placed. This combined with as rigid frame as possible, as well as rubber feet - I hope to minimize it to a level where I will be able to live with it.

I don't have a garage - or a shed for that matter. :P

A full isolation enclosure with external venting would make it a lot better, also take a lot more room (just how big is your desk!?? ). Not sure I'd bother with a chip conveyor/screw, you have enough on your plate as it is. You'll end up using the blow gun, though, with the cover off, and it's going to end up going everywhere. Still, you're already thinking of all this stuff and are one step ahead of most attempting it, you may be able to make it work.