[Zach_G]

I've had this project on the backburner for what seems like forever, just waiting to get all my ducks in a row. Money, time, tools and space always seem to somehow conspire against my efforts but rather than build something with partial dedication, I've waited for the planets to align. Still a few more months to go before I have access to some tooling and can delve into construction but in anticipation I've been hammering out the final design. I'd like to share some ideas and ask questions.


The goal is to end up with a medium sized bench aluminum cutting cnc with some ferrous capability as well. Travels will be X 18" Y 11" Z 13" and rapids of roughly 400 ipm. Spindle is a modified R8 mini mill with a 2 HP treadmill motor. I've already cut down the casting and welded the mounting tabs. Motion control will be with Mach running Zeta4 stepper drives. It's to be seen whether the parallel port will be sufficient or a USB interface of some sort. X and Z steppers are 1200 oz-in nema 34's wired up in a T connection, which will provide performance half that between parallel and series. I got dyno results from Lin Engineering and used them to model the acceleration characteristics of the X and Z. Not too bad for moving 260 lbs! Also assumes the Z axis will be counterbalanced, probably with a pair of gas springs. Will probably limit speeds to 400 ipm and I doubt I can program these acceleration characteristics into mach, and will probably have to settle for a constant acceleration which will result in a linear speed curve with slope equivalent to optimal speed curve at the highest feedrate i want to use. Still should be plenty fast

I don't have any data on the Y motor. Just a round frame nema 34 which I've been told it's around 300 oz-in, but that should be ok for moving the relatively light head along the Y axis.

Linear bearings are 25mm for the X and 15mm for the Y and Z. Plan on ordering precision rolled ballscrews from the Chinese ebay seller linearmotionbearings2008. 20mm for the X and Z, 15 for the Y.

As for the structure of the machine, the bed is an 8x28 replacement bench mill table. The two vertical columns are 2.5" square pre-ground steel bars. the rest of the fabrication is surplus bought 3/4" flat ground stock. I plan on waterjetting this stock on a 5 axis machine to get zero taper on all the reference edges with non critical edges getting a low quality cut to save machining time and cost. The couple local shops I called said this should be feasible but I'd like to know what some of ya'll think. Also would like to know if it's possible to have the hole locations "punchmarked" by the waterjet to be finished on a drillpress later (I don't have a mill anymore and would like to avoid using one as much as possible).

In the end I'd like to have the whole thing mounted on an enclosed bench with flood cooling containment and have the controls mounted under the table.

Well that's probably enough for now, I'll leave the questions to be answered or asked. I'm pretty confident about the design so far, but feel free to fill my head with doubt

[sergizmo]

Zach,

I see a major problem here: inertia. You said the gantry weighs 260 pounds and you want to move it at 400ipm. That is a lot of energy. What does the table weigh? Unless the table is firmly secured to a very sturdy bench or attached to a large piece of steel/granite it will dance all over the place.

If you can get this sorted out it looks to be a very solid machine. Nicely drawn as well. What CAD program did you use?

Good luck, keep us posted,
Serge

[The Blight]

Is there a reason as to why you cant make the gantry stationary, and move the table instead? I know it will fill a bit more, but you wouldnt have to move so much mass around. Like this machine:


I would also make a larger mounting surface for the two columns. Maybe a triangular shape to reduce the momentum arm? I know it will block the path of the Z axis, but if the plate extended behind the column, you would have plenty of room.

I believe I still have a picture of your first design. Has been a great source of inspiration for some of my machine design projects. Never had the time to build any though.

Edit: Beautiful drawings!

[digits]

Very nice drawings, but as others have said, I'd believe you'd be much better off accelerating the table and fixing the gantry - and if you really do have 130+kg of moving mass, you're going to need to anchor the machine to something much,much heavier (250kg+ IMHO).

I'd also suggest that if you can find a way to be able to adjust all the mounting blocks for those linear carriages while the machine is fully assembled, but without its guards, you will have a lot less trouble aligning everything.

Looks like an interesting project - best of luck!

[GITERDUN]

Very nice drawings, but as others have said, I'd believe you'd be much better off accelerating the table and fixing the gantry - and if you really do have 130+kg of moving mass, you're going to need to anchor the machine to something much,much heavier (250kg+ IMHO).

I'd also suggest that if you can find a way to be able to adjust all the mounting blocks for those linear carriages while the machine is fully assembled, but without its guards, you will have a lot less trouble aligning everything.

Looks like an interesting project - best of luck!

My hunch is that we have seen the sum total of this project in those drawings. It seems to me that a benchtop mill you can buy for 1000.00 would do the same job.

[Zach_G]

Thanks everyone (less our Larry the Cable Guy impersonator) for the feedback.

Serge, appreciate calling me out on the inertia. In the quest for bridge ridgidity, I agree a solid footing was overlooked. And since everyone else seems to agree, I can't argue. The best solution I can think of then is... a bigger table. Internet search didn't yield any luck in sourcing just a table. Will call some local machine tool repair shops Monday to see if there might be a used bridgeport 9x42 or equivalent. Any other suggestions appreciated. BTW this was done in Inventor, I also use SolidWORKS occasionally since it has simple CAM/FEA plugins available.

Blight, that's really cool someone recognizes my work! The goal with a moving gantry was to keep things light, since I move every 3-4 years. Good point on the small column attachment. My stock of 3/4 plate dictated material usage and that connection got compromised. I'll redo it.

Digits, I've been keeping assembly and alignment in the back of my head throughout the design, should be doable.

And if Larry can find a bench mill with similar travels for 1k, I'll concede defeat.

Oh, obligatory shot of the pile of parts:

[The Blight]

Lucky!

[GITERDUN]

Thanks everyone (less our Larry the Cable Guy impersonator) for the feedback.


And if Larry can find a bench mill with similar travels for 1k, I'll concede defeat.

Oh, obligatory shot of the pile of parts:

There was a nice Sieg Super X 3 that sold on e-bay for around 1000.00- ball screws, cnc etc.
Forget the treadmill motor- even if you ever got this thing built that thing would last about 24 hours.
I'll mark my calendar to check back in 6 months and see how far you've gone.

[LeeWay]

I don't know about that particular treadmill motor.
I have been using a couple of them. One nearly every day almost all day for almost a couple years now. No issues at all. These are just about the simplest type of motor you can have. Not alot can go wrong. Bearing failure and a shorted wire is about it. Lots of guys have used that same motor without issue.

I think this will be a great project and I look forward to following it's progress.

[CROSSHATCH]

Where can I buy that exact treadmill motor! They no longer have them on Surplus!

I like your design, but like everyone has said. I agree on on having the table move, lots of mass to move around. You will be happier with the quicker table.

Nice project.

-Jason

[Zach_G]

Apparently you can buy it from me in 6 months, but it will only last you 24 hours... ;P

I did manage to snag one of the last few, but similar ones do show up on ebay from time to time.

[chimpera]

I have been working on a similar design. I think that the bridge mill has obvious geometric advantage over the typical c frame. I agree with the others that the weakness is in the leverage at the base of the mill. I think you design can be tweaked to overcome this. too add stiffness to the bridge in the x direction you can move the bearings to the bridges rear and add a triangle support as i think has been suggested. it would also be very helpful to extend the length of the base of the bridge, at the expense of some travel to gain better leverage. the second weakness is bridge flex in the y direction. It will flex like a parallelogram. Without four sets of linear rails at the base you cant count on stiffness in the y direction from the bottom. What you can do is add material on the outside of the bridge so that the stiffness comes from the top of the structure. it should resemble a giant C clamp. If you can stiffen the bridge i am confident your machine will have no problem with steel. I also think 400ipm is more than you will find useful on this size of a machine. if you can achieve 150imp that should be plenty. Good luck, Ben

[riko]

I think this is a very unusual machine with the Z axis mounted on the Y rather than the X. It will be very rigid and have less cantilever than most traditional designs.

Another innovation is the use of net near shape waterjet cutting.

Compare to http://www.5bears.com/cnc.htm which is beautiful work but very labor intensive. And is cantilevered in the Z axis.

Build it as you have designed is my vote.

Don't second guess yourself. Trust your instincts.

[Zach_G]

Hmm, look what I found:

It's about 9x47 and over 3" thick. Seems I need to find some longer rails... It was salvaged from a large machine tool refurbisher about an hour away from me. They had a dozen or so mills just sitting outside going to the scrap heap! I wish I had a chance to look around more and see if there were any whole machines in good shape. Took some effort to get the rust off but it's in decent enough condition. Probably spend the next few weeks updating the design and incorporating y'alls suggestions.

Riko, I love your ideas using a drill press to hack together a cnc!

[riko]

That's a serious piece of metal. What is the weight?

Now that you have a longer Y axis, you may be able to waste some of that for a stiffer machine if you add a little triangular gusset to the Y axis pieces (out the back) so you can add more bolts a little higher up on your bridge uprights.

[CROSSHATCH]

WHat the hell! I would never find this stuff! Awesome !!!


-Jason

[digits]

Wow, that is a long, thin table! Are you going to still go for a moving gantry, do you have the 100" you'd need for a bridge type mill?

[rc-monster mike]

I will throw in my $.02 - that motor won't "cut it". I tried using the same motor on my x2 head a year or so ago - the brushes sparked like crazy and the shaft was about 180F after about 5 minutes (with no load aside from the spindle itself - head was on the table for wiring/testing). Those motors simply don't make enough torque - maybe it could direct drive a Sherline spindle, but it struggled with the x2 spindle. I am using a much larger diameter "treadmill" motor now though - with excellent results.
Aside from that, and the already mentioned z axis attachment at the bottom, it looks super.

[cyclestart]

I was also going to pipe in a suggestion for a fixed gantry.
Then the rescued table appeared.
What does that beast weigh ? What will it weigh loaded with vice/fixture/work ?

You've got yourself the makings of real solid machine. The scrap heaps around here don't have goodies like that table. Plenty of discarded treadmills however.

[GITERDUN]

There was a nice Sieg Super X 3 that sold on e-bay for around 1000.00- ball screws, cnc etc.
Forget the treadmill motor- even if you ever got this thing built that thing would last about 24 hours.
I'll mark my calendar to check back in 6 months and see how far you've gone.

Well 2 of the 6 months have passed- any actual machine progress or are you still wrangling over that 19.00 treadmill motor.