Goodenoughcnc Hybrid gantry

[DKchris]

Hi, a while ago I found this alternative concept to making a gantry setup for cnc:

http://goodenoughcnc.eu/hybrid-cnc/



I really like the straightforwardness of the concept. The square tubes line up everything nice and perfectly square horisontally, and make for nice and wide rolling surfaces for the trolleys.

I was wondering if anyone in here has seen this or perhaps even tried it? Opinions on the construction?


I've already started a build myself, so far planned to be modified in 4 (or maybe 5) ways:

1) frame size will be 900x1200mm - this should supposedly give me a workable area of approx. 665x1000mm, making it possible to utilize most of a 700x1000mm sheet of depron. Nearly all the "depron" I use is 500x1000x6mm Climapor sheets, so the width is less of an issue. And the box table I plan to put under the gantry should also be able to take a 1200x800mm standard sheet of plywood this way, while still keeping the width down a bit, as my workshop space isn't all that huge. I plan to make the table flippable, so the bottom of the box will double as a work surface when the CNC is not in use. Should also help with cleaning the box out .

2) I plan to add an extra X- and Z-axis assemblyy to one of the 900mm end tubes, as it is fairly straightforward to do. This should enable me to do cnc foam cutting as well, locking the Y-axis in place and adding hotwire guides to the ends of respective Z-axes.

3)I'll use proper Angular contact ball bearings for the Z-axis in stead of the proposed standard radial 608ZZ ones. Bit of a no-brainer in my opinion, since I can get those bearings easily locally - maybe not something everyone can....the apparently "everywhere-obtainable" 608ZZ's will properly do great for all the rest.

4) I'm using 40x40x2mm AISI316 stainless for the frame in stead of the spec'ed 40x40x3mm ordinary steel ones, so I won't have to keep the frame lubed or coated to avoid it rusting (In Denmark everyone practically lives on the ocean shore, at least with regard to salt in the air ), and to save a little weight. I haven't really done the math, but I'm betting on the higher "stiffness(?)" of the stainless to about even out the reduction in wall thickness.....also, it cuts the price of the stainless a bit.

5) I am also contemplating lengthening the 2 Z-axes a bit; I am considering ~100mm extra, giving me about 250mm of movement with the Z-axis assembly total height growing to 400mm. Partly because I already have the materials to do so at hand, so it won't really cost me any extra, partly because of my wish to incorporate the cnc hotwire ability. I am not all that worried with regard to the accuracy for depron or hotwire cutting, but I am a little concerned about the impact on accuracy for milling and 3D-printing. Any inputs on this?

My progress so far is:
- Sorting out drawings and build instructions, parts availability and cost.
- I've obtained, cut and drilled the steel tubes for the frame, I have cut out all the other steel parts for the trolleys(100x100x3x130mm steel tubes) and belt tighteners(from the leftover stainless tubing).
Next will be marking and drilling the 4 trolley tubes, which is probably going to be the most critical part.

 

[DKchris]

Progress:


Lined up the mechanical setup and the electrical parts for an overview.(Why do I always feel like one of those "Here's what I ate today" before/after Instagram post women when I do this??)

As mentioned the tube parts are ready for use, trolley and belt tighteners need drilling, currently done marking up the parts for this. (I don't currently own a drillpress myself, so waiting for opportune moment to borrow one.)

Left in cnc frame: Bag of mechanical tidbits like t-nuts, flex couplings etc., 6 polulu 8825 2.5A drivers and a wiring box with an emergency stop button.

Right in cnc frame: 2 MDLCNC Parallel port breakout boards with cables, Motors (4xNema23 110Ncm, 2xNema17 48Ncm). Motors are all 3-3.5V rated, so I intend to run everything on an old 12V/46A Server supply I've got; Ought to work out, and it also delivers 5V/5A which should suffice to supply the signalling bits.
I'll build 6 little carrier circuit boards for the polulu drivers using proto board and install them at each of the motors. They'll each have a reasonably sized supply storage capacitor and a couple of small ones for electrical noise mitigation, a 3A pin thermofuse in socket, a dip switch for the microstep settings and proper fasteners for all cables and leads.....and a socket for easily exchanging the polulu, off course. I'll probably also prepare them to carry a small 40x40mm12V Fan each, if the polulus end up needing extra cooling for their tiny heatsinks. I've already got all the parts available.

Front Right: The top of the 30x60 T-slotted alu profile I intend to cut up for the 2 Z-axes. It's currently used as the central carrier frame of my rather rarely used gravity driven hotwire setup. Neat solution, by the way, and easy to take apart again.

In the red box back right: A cheap (Think Harbor Freight) 1000W wood router a friend gave me for parts a long time ago, since I had one like it myself; I believe I paid about the equivalent of 50$ for mine (Including 25% VAT, mind you). He had given up on his as the power switch had given in; he had replaced it with a switch on the lead and shorted the internal one, but it had started to make "unpleasant screechy noises" after he had put it back together. Other than that it ran OK with variable speed and everything when I plugged it in back then. Put it on the shelf and forgot about it.
Got it out and took it apart the other day to have a look; was surprisingly easy. Turned out the noises came from a poorly finished "push-fit" plastic separator part inserted between the motor and fan. It had a lightly bent edge, which had gotten caught somewhere on an edge of the motor casing and been pushed down so it just touched the fan part. Edge was cleaned up and everything put back together correctly, and now the unit worked just fine, no unintended noises, and quite nicely balanced too. Decided to take it apart again and try to figure out if I could "clean it up" by removing everything not neccesary, and then make it mountable on the 30x60 profile. Would make a real cheap "first experiments" router unit for my CNC gantry if possible. And it looks like it will be. Isn't even all that heavy, it's just under 3kg (~6.5lbs) when fully assembled, and my guess is at least half of that disappears with the base/Handles/sliding mechanism/fence etc. A mount will add some weight, but a similar Kress unit weighs about 1.7kg(3.75lbs) and would also need a mount, so i'm fairly positive about the whole thing.


Apart from drilling holes in the mentioned parts, all I need to get before assembly can commence will be toothed belts and pulleys, and most of the nuts and bolts.

 

[rockyboy]

I like that this design allows you to (carefully) drop the machine on top of a work piece in place - such as routing inlay designs into already installed flooring or large tables.

 

[DKchris]

If you adjust leg length on a MPCNC and weight the feet down to hold it in place, it'll probably do the same.

I see this as a fairly direct alternative to the MPCNC for those who do not have access to doing 150hrs of 3D printing up front, an where sourcing printed (and other parts) from the US is prohibitively expensive.


If you were to fit it with 6 Nema 17's and could live with the limited speed (This is a bit heavier than the MPCNC, so inertia will limit max. speed if you want to stay in control/accurate. On the upside it ought to be more rigid), and know where to source the steel tubing locally(around here the freight is typically 8-10 times the steel cost for the parts needed), it'd probably end up at a sort of similar price tag "over there".

My own cost estimates:
I got the 40x40mm and 100x100mm tubing for the equivalent of about 75$ total. This was fairly cheap, considering the stainless tubing.; I had the luck to find some good people. Ordinary steel tubing would have been fairly easy to find at this price.
Nuts/bolts, electronics, fittings etc. ought to amount to about the same, given the same control solution is used.

I've actually calculated my construction/combination to run into the equivalent of about 600$ for the basic 3 axis setup, and about 700$ with my 2 extra axes for hotwiring (With 4 Nema23's, mind you). And nearly everything is sourced fairly locally, so with 25% VAT(steppers and MDLCNC boards are from Letmathe in Germany, so 19% VAT; Came at a very affordable price though). Should very likely be possible to get a lot of the bits and pieces a bit cheaper on the US side of the pond. I did have to put in the time to find and buy everything myself though. And some stuff like the PSU and the aluminum bits I already had, so they are not included in the price.
I've chosen to start out with LinuxCNC/parallel port control, just cuz' it's a lot simpler to setup for the first bunch of experiments with lining everything up, adding different tools etc. I can live with the speed disadvantage and have a PC to dedicate for it. To make the 3-axis version stand-alone like the goodenoughCNC and MPCNC kit versions I'd have to add another 35-40$ for an Arduino Mega and a Ramps board; I'd have saved about 20$ on an MDLCNC board, so not really a big difference. Might very well be added later.

The Goodenoughcnc.eu complete kit price for the 3-axis setup was around 800$ plus freight, not including the 40x40mm steel tubing; Would probably have ended up around 1000$ total when finished. This did admittedly include their special TOSlink optical fibre controlled driver solutions for all axes, making the setup very immune to arcing and other electrical disturbances, so directly usable with a plasma cutter or perhaps even a MIG welder if you would want to dabble a bit in metal 3D printing. And also capable of handling 5A of stepper current. A bit pricier, but not really a bad offer to be honest. Sadly they stopped shipping kits before I started my build.

Vicious CNC state the MPCNC to run into 400$ if you have a 3D printer(and know how to use it ), ~500$ if not. The full kit with all printed parts and everything needed for assembly (except for the 25mm conduit tubing, ~50$ total over here for the standard 18') sourced from vicious CNC would be about 460$ plus another ~100$ for freight to Denmark; then danish customs would add another ~25$ in handling fee and finally calculate and add 25% VAT to the total, so ~730$ plus the 50$ for the tubing(~100$ for tubing if I would want it the same size as my Goodenoughcnc rig).

So cost would have been in a fairly similar range to me either way.

 

[ChrisV]

I've been scouring the internet for hours in search of references to the GoodEnoughCNC mill, and you're just about the only person I can find that is actually building the thing .
How is your build going? Any exciting updates?
I've been wanting to build this mill for about half a year now, but I've been stuck researching the different parts and weighing all my options:
1) I was personally thinking of building the entire gantry in aluminium. There's a company in Belgium that can cut all the tubing and plates to size. If I only have to drill the appropriate holes, the build should be a whole lot easier (especially since machining aluminium is a lot easier than steel). I guess it will also put a lot less strain on my stepper motors since everything will be lighter and the added benefit is that it won't rust.
2) I also considered increasing the z-axis size. Since this is my first build, I do have some concerns that the increase will make the z-axis less stable (more wobbly).
3) I don't have ready access to a 3D printer, so I'm looking for some alternatives to the 3D printed parts. I found some aluminium clamps for timing belts that I could use instead of the 3D printed ones. I'm also hoping to replace the linear guide mounts by these mounts and try to replace all the remaining printed parts by wood and/or custom ordered parts.
4) I'm hoping to replace all of the controller electronics with a Protoneer Raspberry Pi CNC hat together with some external TB6560 drivers (although I might make do with some pololu drivers)

I still have some doubts before I start construction though. I want to use this platform for milling aluminium. I don't expect any high speeds or stuff like that, but would a belt driven cnc machine like this be sturdy enough for milling metals?
The problem I'm having with this machine is also the fact that the bill of materials is seriously vague. It doesn't suggest an exact type of NEMA stepper motors, nor does it give explicit details about the timing belts and pulleys you should use.
I stumbled upon the following timing belt and pulley. Any suggestions on whether or not these will suffice?
I hope your build is going well. It might be nice to be able to talk to someone else about how to tackle the problems with this build .

 

[DKchris]

I've been scouring the internet for hours in search of references to the GoodEnoughCNC mill, and you're just about the only person I can find that is actually building the thing .
How is your build going? Any exciting updates?
I've been wanting to build this mill for about half a year now, but I've been stuck researching the different parts and weighing all my options:
1) I was personally thinking of building the entire gantry in aluminium. There's a company in Belgium that can cut all the tubing and plates to size. If I only have to drill the appropriate holes, the build should be a whole lot easier (especially since machining aluminium is a lot easier than steel). I guess it will also put a lot less strain on my stepper motors since everything will be lighter and the added benefit is that it won't rust.
2) I also considered increasing the z-axis size. Since this is my first build, I do have some concerns that the increase will make the z-axis less stable (more wobbly).
3) I don't have ready access to a 3D printer, so I'm looking for some alternatives to the 3D printed parts. I found some aluminium clamps for timing belts that I could use instead of the 3D printed ones. I'm also hoping to replace the linear guide mounts by these mounts and try to replace all the remaining printed parts by wood and/or custom ordered parts.
4) I'm hoping to replace all of the controller electronics with a Protoneer Raspberry Pi CNC hat together with some external TB6560 drivers (although I might make do with some pololu drivers)

I still have some doubts before I start construction though. I want to use this platform for milling aluminium. I don't expect any high speeds or stuff like that, but would a belt driven cnc machine like this be sturdy enough for milling metals?
The problem I'm having with this machine is also the fact that the bill of materials is seriously vague. It doesn't suggest an exact type of NEMA stepper motors, nor does it give explicit details about the timing belts and pulleys you should use.
I stumbled upon the following timing belt and pulley. Any suggestions on whether or not these will suffice?
I hope your build is going well. It might be nice to be able to talk to someone else about how to tackle the problems with this build .

Hi ChrisV

Apart from recieving the last of the neccesary Hardware parts (Bearings, belts, pulleys etc.), I have nothing new to report; I've had to "short term shelf" the project over the summer holidays and concentrate on having a good time with the family (not the worst of trades) and have been a bit slow in getting back into it; Hence also the late response.

My 2 (euro)cents:
On 1) Aluminum will probably work for hobbyist use, even though you'll probably have to go up quite a bit in wall thickness to obtain the same rigidity. One thing to note is that you have to give some thought to what materials you combine. Zink and alu don't go well together and will cause galvanic corrosion of the aluminum. So avoid galvanized bolts etc. I'm honestly not a buff on this, but I remember having to opt for A4/316 over A2/304 stainless bolts on my Motorcycle motors, as they were less likely to generate corrosion with the alu engine case, cylinder/top block and covers, as they are made from aluminum alloys, that are more prone to generate galvanic current with other metals(but might be more rigid or better for casting?). The hitch is that different aluminum alloys react differently with a given other metal. Has to do with where the alloy and the other metal are located in the reactivity/galvanic series. Found this link for some info: http://www.preservationscience.com/materials/metals/PGC.html

On 2) I still have similar concerns; But the thing is that the only parts that needs to be longer for a longer Z axis is the threaded rod, the round steel rod for the slider bars and the alu profile(s). For all, I'll have to buy (or allready have)them in meter lengths anyway, so It's just a matter of cutting less material away for me. And if it should prove too unstable after all, I can just unbolt the lower end slider bar mounts, run the entire Z-axis bit up and out of the gantry, disconnect the stepper, dismantle the Z-axis(real easy) and cut the 4 parts to length for a new shorter Z-axis. No big hassle.

On 3) Sounds OK to me. Maybe consider some flat rail linear bearings? The Goodenoughcnc guys posted an updated version , I've forgotten where(was it in the Github drawing source??), that put the 2 alu profiles on either side of the stepper/threaded rod with such linear bearings between the Z-axis carrier plate and the back side of the profiles, and a modified top and bottom plate between the 2 profiles (These should be simple to guestimate) for improved rigidity and generally better performance for the Z-axis. I don't have 2 separate alu profiles, so I'll stick with the original design.
The standard guide rods could also be mounted simply by drilling a (6mm?) hole through each end of the rods and putting a bolt though it for clamping the rod directly to the alu profiles. Might need to grind a pair of flat spots at either end of the drilled holes for bolt head contact face and to clear the clamps that goes in the profile slots, but it would still be rather simple to do.

On 4) Makes sense. I went with the polulu's to keep it cheap for starters, with the intent to upgrade later if I found out I wanted to dig into the full capability of the Nema23's. I still wen't with the 23's though, since the specific offer I happened to get on them was not much more than the 17's (like 23 vs. 15 &#8364......and off course because I wouldn't have to redesign the mounting holes for them. Bit lazy, I guess.




I wouldn't expect my setup to handle milling even mild steel, allthough very thin plate might be possible....no real clue though. The original allegedly handles routing dense plywood quite well, so I'd imagine it could also be coaxed into milling not too thick sheets of soft alloys.....with the right cutting tools and a slow advance speed - I wouldn't expect anything even remotely near 100th of a mm type of accuracy though. A thick block of alloy? Honestly....no clue. We need to reach out to any experienced machinists out there for that one. Or just build the darn thing and try it.

On cutting Alu with a rig based on Alu rails........to be honest I dont really find it makes sense. My gut feeling says you need tools that are more rigid than the material you are trying to cut. But I might be wrong. I have no experience or math to base my guess on. All I really know is that the steel version should be "more" rigid than the MPCNC.....and the MPCNC can do all the main things I want my rig to do......and that's basically what I'm going with.


On belts and pulleys I've had similar problems finding especially the right size belts in 1.5m lengths without having to purchase an entire roll of "endless" belt, costing a small fortune. So I have actually gotten the exact same parts as the ones you link to, albeit from other suppliers. Main reasons are: They were fairly low cost (about 20$ for a 5m length of belt and about 5.50$ each for the pulleys at a local supplier), they are the exact type of belt and very similar type pulley to that used in the MPCNC(they use 16T pulleys with 5mm hole for Nema17's), so the combination should not be THAT shabby....and a wish to get it going ASAP. These belts are definitely too flimsy for using the full potential of the Nema23's though. They will probably not break, but I expect them to be too flexible to get proper accuracy when routing more substantial materials, so I expect to upgrade them to the 10mm width T5 belts spec'd in the BOM later when/if I find a suitable supplier. And they will most likely have no issues handling Hotwire cutting and foam plate cutting with the needle cutter, that are by far my first and highest priorities for this rig.