Has anyone built, or seen plans/feedback, on building a small footprint 12x12" for milling some D2 or M4 tool steel. The tolerances don't need to be that great, but I need it to do some decent 3d modeling. I looked at the C-Beam and think that is pretty close, but I want linear rails. Was thinking of building one myself, but wanted to see if anyone had feedback on this. Curious what folks think about moving to a steel frame instead of aluminum extrusions. Would this give the system more rigidity? Is this even possible? Thanks in advance for any feedback.
I don't think it will be possible, not with aluminum framed machines, what you want is a proper milling machine.. steel uses low speed cutters and flood cooling, not something you can do on these machines. I would consider getting a grizzly or similar small milling machine, and a cnc conversion kit.. it will be more expensive (by like double at least), but none of the machines here, even the custom ones with upgraded parts, at least as far as I have seen, will be able to cut or mill steel.
go to tormach.com and pick one, aluminum framed machines are not for cutting steel (it is possible, just not a good solution at all)
Heh.. straight to the pricey ones huh cheapest one they have starts at like $5500... It's why I suggested a Grizzly, or maybe a Precision Mathews PM25 with a cnc kit.. total cost will probably run 3k.
oh, Tormachs are not pricey (-: what we all really want is this TM-1 | 40-Taper Mill | Toolroom Mills | Vertical Mills – Haas CNC Machines
Speak for yourself, I want a Speedio. (Though I guess I'd settle for a VF2SS) My G0758 can mill steel, I just got through four steel end plates made from 1x4" 1018 bar and dimensions came out fantastically. I wouldn't recommend duplicating mine, though, it's too small and fiddly; I'd get a G0704 (minimum, maybe a G0795 or G0761, something in the $2.5k range) and put 30mm rails and double-nut 20mm ballscrews on it. You'd end up with something that meets or exceeds the 1100M's performance at closer to the 440's price. Converting larger machines is way easier, there's space to do everything. That said, you're gonna be lucky to find 12" of Y travel at any kind of reasonable pricing. If your project can fit in 7-9", you're doing ok. If it has to be 12" of Y... I'm not sure. Even an 1100 only has 11" of travel at around $25k to get it actually cutting. A G0720 has like, 9.5". I'd probably look for cheap used machines. You can probably find, like, a 2007 low-end Haas- TM-1, VF2, whatever- in ok-ish condition for $15k in the right area. That would be more in a non-manufacturing area, and you'd still have to pay for riggers to move it, but it'd be the best way in. Bear in mind, on any machine, spindle speed though; it's the spindle that's gonna be the constraining element. A 2,000rpm spindle (like a Grizzly R8) is slow, especially with sub-1/4" tooling. You're talking single-digit IPM, 100-200mm/min, down to maybe 2IPM for 1/16" tooling. Mid-range spindles, 6-12,000rpm, are where machines tend to be happy in steel. You can buy cartridges for a conversion for $500-1000 (depending on speed). High-speed spindles (18,000-40,000) are good for a lot of small tooling, but if they have no low-end torque they won't work for roughing ops. Fusion 360 defaults a 1/4" 4FL to around 8,000rpm, so your machine still needs grunt, unlike a router. What you could do is buy a manual mill for $2.5k, and use it to build a machine from scratch directly, say, onto a surface plate. That's a semi-common option and wouldn't be all that expensive but you would need to be quite proficient at machine design and metrology. But because C-frame machines inherently have a column in the way of the Y travel, you're always gonna be constrained in that direction without buying vastly oversize.
I have a tormach. For what it is I love it. I see them come up for sale all the time on craiglist. But the same can be said for the TMs that David suggested. We used them in school and I really liked them. I would have bought a used one, but I didnt know anything about the mechanics back then. If I were to buy a used mill today, I lean toward the Haas.
I'd get a used VMC over a new Tormach any day of the week, not even a question. Having now converted a mill, I just don't see their value proposition to the mechanically-inclined any more. A VMC is a very different beast.
Thanks very much for the help! It really helps a lot. I misstated my need for Y dimensions, could easily get by with 12x4 or so. I do have some experience on a Tormach 1100 so am familiar with that mill. What about building a single purpose small machine based on a steel frame? Easier to automate a larger mill that build from scratch? I think it may be easier to just out source our current needs but I wanted to check what it would take to do it ourselves. Thanks again!
My G0758 (Grizzly G0758 Benchtop Mill Conversion) has about a 5x13" travel. Any medium-size benchtop mill will get you there for those dimensions. That's absolutely, unequivocally, the fastest, cheapest and best way to get to a point to cut steel accurately. Unless you happen to have in-house or nearby 24" surface grinders and heat treat ovens for weldments, I suppose. Depends what the product is, too. Single purpose machines have their place, not everything needs to be CNC, but you said "3D modeling", which I interpret as "3 axis surfacing", which definitely needs CNC.
I have two basic needs right now, a single very simple 3d pocketing in aluminum. Doing this on my aluminum CNC now but can "see" the machine flex. Overcome this by running a cleanup pass at the end but it would be nice to get it all done once. Plus, the mess with the chips is less than desirable. The second is a new venture and basically involves creating a knife which is a single piece of steel, no traditional handle, and needs a 3d CNC path in order to automate. The knife itself is about 10x3. I am pursuring options to create a single purpose solution where we can just pop in a block of metal, machine it, flip it, machine it and have a knife that needs heat treatment and blade grinding.
Similar to that Diresta skeleton knife, and (perhaps moreso) the short-lived Grimsmo Tor? Though 10x3 suggests more of a chopper or even small cleaver/chef slicer. In any case, after generating screw comp files through LinuxCNC, my mill appears so far to hold a thou easily, often more like 3-4 tenths, in aluminum, and probably not much more than that in steel- I'm VERY happy with the fits I've gotten from these plates I just ran. I'm running Saunders fixture plate and Mod Vise on it, no issues with flipping. Made pallets with Mitee-Bites for dedicated bar stock work, also extremely accurate. Generally pretty happy with its performance. It also has a quite flexible head and column, and is only running low-grade SFU1204 and SFU1605 screws on the X and Y respectively. The Z axis performance has increased significantly since moving to a double-nut DFU1605 screw, which would be easy to fit in all axes of a far-more-rigid G0704 or similar (I think there are even off-the-shelf conversion kits for that particular one). I'm running closed-loop NEMA34s, about 4.5Nm on X and Y and 12Nm on Z. The mill originally cost about $1000, all-in I'm probably still sub-$3k, even after the $600 on the motor/driver kits. The $1100+ on the 12k BT30 spindle cartridge and air cylinder will bump that up significantly once I install it instead of looking at it in a toolbox drawer, of course, but that head will move to a future, larger machine (more like the ones I described above, unless I manage to make so much money from this G0758 that I can go straight from it to a VF2 or whatever) where its power and rigidity will be of more use. I think I was pretty thorough in image documentation in that thread. If that seems like something you can do, then go for it. Otherwise, a P440 will probably get you there but cost twice as much.
Good luck with that.. ideally needs 440V... not something you can run in your house.. or for that matter.. most non industrial zones.... Also.. 30k is still not that bad.. but they are completely different classes.. one is major manufacturer.. the other is small business or lab. But the Tormach is on the high side for hobbyist.....
Here is the problem with some of those used machines.. replacement parts.. any used machine you get more than likely will be heavily used.. which brand will have the cheapest replacement parts.. and which will require you to mortgage your house...
If you're running 440V, you have to get the HV transformer, Haases run on 220V. TM-1 will happily run on a 40 or 50A 240V 1PH stove line (though I think non-students would rapidly outgrow it). The Mini Mill will both (barely) run on a 30A dryer line (no heavy cuts) AND fit under an 8ft garage door. Haas absolutely makes hybrid hobbyist/micro-manufacturing machines, with legit capabilities, for not too much more of a stretch than Tormach's prices. They also carry a full stock of spare parts and release the documentation for users to do their own maintenance, no HFO required. They're also by far the most popular machine tool in the US, and so comfortably have the lowest-cost replacement parts and resale prices. (I'm not a Haas fanboy, they're just priced to be looked at a lot, and you tend to hear about them incessantly on the BoM podcast...) Tormach sells machines for twice what they're worth. For people who need in-house manufacturing but lack the expertise to put together their own machine, it's almost certainly still a value- most people's Tormachs have paid for themselves many times over, I'm quite sure. I just don't find a value in them personally because at this point I can put together something near-identical for a third of the cost in a relatively short amount of time.
I don't find value in any of them..because I cannot afford any of them . This guy from Germany (https://www.youtube.com/channel/UCJK1WpPzUjz2uIRahllinvw/videos) has made an awesome Epoxy Granite DIY mill that honestly, and this guys from the Netherlands made an even better one.... I think they can probably rival most of the sub $20k machines... although they don't say what their costs were.. so it may not necessarily be a cheaper solution....
