New and in need of wisdom

[Washout]

Hi all,

Been lurking for a while and as I'm about to get an extension to the man cave, I'll be filling that with a new toy.

I've been making things out of aluminium and mild steel mostly (some stainless as well when needed) for a variety of projects for a number of years, but only having basic tools and a bench drill is limiting what I can achieve (especially in terms of accuracy). So I'm looking to buy or make a CNC mill, which I also want to double up as a 3d printer by swapping out the spindle with an ABS extruder and tossing a heated plate on the bed. I've also been 3d modelling/CAD'ing for a number of years (some PC sim/games related) and use a variety of software dependant on what I want to do and how quickly e.g. AutoCAD, Blender, Sketchup, 3DMax and some proprietry stuff.

Ideally I'd like a 600-1000mm by 400-500mm bed and a fair bit of Z travel for sculpting/3d printing work and I will likely be getting a 4th axis in a few months after initial purchase. Work will be predominantly in Aluminium (6061 T4/6 mainly but may need some harder/stronger) with occasional mild steel and some ABS and other plastics.

The options I have been toying with after reading here for a while are:

1. Buying a ready made CNC mill from the UK - expensive and from what I've seen of machines like the Heiz etc they are generally not using supported rails and some other components I read on here should be almost mandatory.
2. Buy a cheap chinese ebay machine such as the 6040 - would buy in the knowledge it will need some or most of its electronics replacing and supported rails in the Y and Z axis.
3. Design and build my own machine from scratch - longer lead time and some additional digging around for components, although thanks to some of the build logs I have already got most sourced e.g. extrusions, rails, ball screws etc.

I have almost discounted option 1, as I haven't seen anything that doesn't look like I'm tossing 100% mark up at the builder and I'm the kind of person that fixes things myself rather than enduring the frustrations of after market support.

Option 3 is appealling, but my main concern is I don't have the kind of tools that I think will achieve the kind of accuracy I think I need to avoid things like binding on rails/ballscrews due to misalignment from imprecise drilling. I also don't have welding kit and have very limited experience in this.

Option 2 is looking favourite at the moment therefore and even if I double the cost by upgrading then I'll still be around my budget.

Just in case I'm worrying about option 2 needlessly I have the following design I whipped up over a couple of hours in Sketchup, which if anyone has the time would like feedback on (most of the extrusions are 30mm or derivatives of) and the bed is 1000mm by 500mm to give an idea of scale but happy to send anyone the .skp if they want to dig further/look around:

Thanks in advance.

[blackburn mark]

its almost there
looks a bit tall.... the way i see it, what you gain in wraping your gantry around your bed you will lose on the extra length of your gantry sides (as far as rigidity is concerned)

find a way to keep the gantry side lengths (hight) to an absolute minimum

your X axis linear bearings are a tad close together
off the top off my head i would say you need to be coser to having your X axis bearing half as far apart as your gantry is tall and if not as far apart as you can (if that makes any sense) especially if your looking to cut alli

im pretty sure you will not cut stainless with much joy on any router type machine

[Washout]

Thanks Mark,

Yes it probably could do with a bit of shrinking vertically - easy enough to do in the virtual world :)

By the X Axis did you mean the two running left to right on the gantry? (I thought that was Y...oops) - and if yes I was also concerned about that due to posssible twisting movement being possible - I can add a new extrusion below the existing 120x30mm and move the lower rail to that - should also give more room for the ballnut and Z axis platform coupling/mount.

Copy that also on the stainless thing - I only do a bit of cutting ground shaft now and then in stainless for mounting gears and the like on, so no need to mill stuff in stainless, but I guess stainless can be done on a tougher rig? Ali is what I will mainly be working in as I said, and would hopefully like to cut 7075 (again not my primary material) - I'm not that worried about speed, as the main aim is to get shapes I just can't with hand tools/pillar drill.

Thanks for the feedback.

[Jonathan]

You will need a really strong machine to cut steel. If you can afford a milling machine that would be the best option for steel, unless the parts are quite large?

It looks like you're planning on having only one ballscrew on the X-axis? The machine will be a lot stronger if you use two, one on each side of the gantry.

[Washout]

Thanks Jonathan

I have been thinking about a second ballscrew on X, having read a few posts. I understand a timing belt arrangement and tensioner of some description is the way to go or should it be chain driven?

If I can squeeze it in budget I'll put that on, as the gantry isn't going to be light I expect and spreading/equalising the load on it sounds sensible.

The steel work would be small parts and yeah probably for a different machine.

BTW - I've been looking at mills, trying to find something suitable to cut "dovetails" into an aluminium recess that a bevel edged 3mm plate can slot into - any idea what I would need to do that?

Cheers



Chris

[irving2008]

...
BTW - I've been looking at mills, trying to find something suitable to cut "dovetails" into an aluminium recess that a bevel edged 3mm plate can slot into - any idea what I would need to do that?

Cheers

Chris
You need a dovetail cutter, but depends how wide/deep the slot is as to what size to buy. Something like this: http://www.ebay.co.uk/itm/16mm-x-45-Degree-HSS-Dovetail-Cutter-End-Mill-/120938840568

[Washout]

Perfect irving2008 - exactly what I need - the main block of ali will have a recess made into it 3mm deep and the dovetail will be around the edge of that - the chamfered plate will then slide into that so it is flush to the block surface. That mill will also solve another problem I had which is how to cut adjustment slots for a latch which will be mounted into the sliding plate (the plate itself will have to be a milled shape to make that work). Quick and dirty sketchup to show what I'm on about and thanks for the link:

[JAZZCNC]

Chris from the looks of it you really do need a Mill.! . . . Router style machines like what you have started designing don't really do cutting steel and aluminium so I'd forget that design.!
Router/(Weak)Mill can be done but needs a very different design. Steel and aluminium need a very strong dense machine which will absorb the vibrations.

You also have a slight conflict regards the materials and the spindle speeds they require for cutting.? Cutting steel requires low spindle RPM, this is also quite low for certain tools like dove tail cutters, which require sub 1000 RPM. Aluminium requires much higher speeds for best finish. The difference between the two can and will be several thousand RPM and while Aluminium obviously does cut at lower RPM's it much prefers higher speeds to get best finish and cutting feed rates.
If you plan on cutting high end Aluminium like 7075 then I suspect finish and appearance will become important so the relationship between feed rates, spindle speeds and machine strength plays a big part.

So the point to me saying all this is that you need a very strong machine and a mill will probably provide this best. . BUT . . due to your requirements of cutting 7075 type aluminium mixed with steel using cutters like dove tails then it will need to be one with a very good spindle providing wide speed range.? . . . These are few and far between and usually come at a cost.!!

Don't let me put you off building a router style machine and even with limited tools it's very possible to build a great machine. . . BUT I WILL . .try to put you off building one with the intention of cutting steel or high end aluminium unless very specific designed to do so. ~(I will also advise the cheap Chinese machines are not up-to the job either so save your money.! )

Sorry if this may not be what your wanting to hear but it's true and cutting steels or aluminium to a good standard requires a very different league of machine and unfortunately this does come at a high price compared to standard router type machine.!

Save your money and buy a Mill with decent spindle is my advice.!!. . . OR. . think long and hard before building a machine looking at a very different style machine than what you have been.

Good luck.

[Washout]

Thanks Jazz,

Plenty to think on in your response, so just for clarity:

It would seem that I need at least 2 machines to cover the range of materials I work with: CNC router for softer materials up to say 6061 T6 (will a router style rig handle this grade?) and a Mill for 7075/Steel work.

I had been looking at the Sieg X3 and some other similar machines a while back - I'll revisit those and see what's available with CNC control, either out of the box or after market.

I'd still like to get a router style machine running, for softer materials, so will still pursue that avenue for build/purchase initially with eyes duly opened.

Can you clarify or help with a few things:

When you say "stronger" would this mean replacing many of the extrusions (especially on the gantry) with aluminium plates, upping the size of the ballscrews and supported rails?

I have already lowered the Z axis, widened the spacing on the Y axis supported rails and added in two ballscrews on the X axis - any other areas I should look at and should I look at flood cooling from the outset?

BTW - I don't mind building a purely extrusion based machine first and then using that to cut out the plate gantry sides etc to make the machine more rigid later (I'm sure I can use the "wasted" extrusions I replace for other projects).

A couple of other areas I could also do with some advice on are:

Software - CAD isn't a problem, so no worries there. CAM/toolpath software I have seen is the Vetric stuff and VCarve Pro appears to be what I need (I've also been playing with Cut3D already just to try stuff out) the and Mach 3 for the machine control? Any others out there?

Oh and cutting carbon/glass re-inforced resins and plastics - can this be done, as I also do some composite work?

Apologies for the flood of questions, I'd rather get at least some of the direction of travel sorted now, as I won't be ordering anything for a month or so, as there is much to do to make the space for the new tool(s).

Thanks again


Chris

[JAZZCNC]

It would seem that I need at least 2 machines to cover the range of materials I work with: CNC router for softer materials up to say 6061 T6 (will a router style rig handle this grade?) and a Mill for 7075/Steel work.

Yes and NO.? If steel then YES you really need a milling machine. . . .BUT. . . If you choose a different design router style that doesn't use long gantry sides but instead has strong frame with high sides which the X axis rails sit directly onto so gantry pretty much sits directly on the bearings so reducing flex and handling resonance's better. Along with driving the gantry from both sides with twin ballscrews then coupled with raising or lowering the material near the cutter either using a adjustable bed type system or just putting in tempory raising blocks you'll be able to cut a vary good range of materials right upto 7075 aluminium. Obviously you'll still need a decent spindle for this a run of the mill wood router won't cut it.

You could also build a fixed gantry style machine which tend to be slightly stronger design.? It just requires more working space for the same cutting area.

Search the forum and you'll find loads of references from Jonathan and me regards this type build has it's one of the few things we agree on.!!


When you say "stronger" would this mean replacing many of the extrusions (especially on the gantry) with aluminium plates, upping the size of the ballscrews and supported rails?


I have already lowered the Z axis, widened the spacing on the Y axis supported rails and added in two ballscrews on the X axis - any other areas I should look at and should I look at flood cooling from the outset?

Has I suggested above it would be a completely different design really but would or could still involve profile and alu plates. Regards upping size of ball screws and rails then no that's not required and can actually be detrimental to machines performance if too big.

Flood cooling is always good idea but when cutting Aluminium then blown air works very good, esp if used with a misting system.



Software - CAD isn't a problem, so no worries there. CAM/toolpath software I have seen is the Vetric stuff and VCarve Pro appears to be what I need (I've also been playing with Cut3D already just to try stuff out) the and Mach 3 for the machine control? Any others out there?

Oh and cutting carbon/glass re-inforced resins and plastics - can this be done, as I also do some composite work?

All those you suggest are OK for wood and v-carving, engraving type work but you may want to look at others that are more suited to engineering type.
The difference comes mainly from the 2D & 3D tool-paths they provide which tend to be more suited to milling metals. Things like Drilling, boring, chamfer, thread milling cycles. They also tend to come has standard with more material entry options, things like ramping, contour ramping, spiral, parallel roughing etc.
3D tool-paths tend to come with features like advanced roughing, pencil, flatlands, equidistant offsets etc

Bobcad cam, Dolphin, Mesh cam and even Cambam I believe offer some of these options and his quite cheap.
Personally I use Bobcad cam for most of my basic mechanical metal type work and Artcam for the more Art'y stuff like Engraving, V-carving and general wood/plastic type work.
Really there hisn't one software that does all aspects best so you may find your self using a few different software's depending on what your doing.?


Apologies for the flood of questions,

Don't Apologise and keep asking untill your 100% happy and understand. Esp ask before buying components if your not 100% sure.

[Jonathan]

Steel frame is almost certainly the way to go mainly because it's cheaper if you buy it in the standard 7.5m length. Don't be afraid of welding!


Software - CAD isn't a problem, so no worries there. CAM/toolpath software I have seen is the Vetric stuff and VCarve Pro appears to be what I need (I've also been playing with Cut3D already just to try stuff out) the and Mach 3 for the machine control? Any others out there?

Since you've already used 3DSmax (http://www.cnc-toolkit.com/), you shouldn't find it too difficult to use CNCToolkit which is a free script for 3DS that creates toolpaths up to 6-axis. It's not the easiest CAM program (if you can call it a program) to use, but I think that's mainly due to how different 3DS is to most CAD programs. Having said that if you only need to to 2.5D milling, then there's no point overcomplicating the process - just use CamBam, one of Vectric's programs or similar.


Oh and cutting carbon/glass re-inforced resins and plastics - can this be done, as I also do some composite work?

No problem, with the right tool:

http://www.mycncuk.com/forums/showthread.php/4408-Cutting-Carbon-Fibre?p=29541&viewfull=1#post29541

[Washout]

Thanks Guys,

Here's where I got to before your latest comments came in - also threw some nema23's and a spindle on there just for show and the x axis one will likely go on the corner with a second ballscrew on the other corner linked by belt or chain.

Maybe a silly question, but is the tool chatter etc linked to harmonic resonance in the frame essentially and if so would assymetric mass damping help i.e. attaching weights to the extrusions in irregular places to stop it singing like a tuning fork?

[Washout]

Gentlemen,

Another configuration for your consideration - I think is what you guys were on about getting the gantry sides "shorter" by putting the Y axis on top of the x axis bearings. I guess the trade off will be the Z axis being longer below the y axis rails?

Will probably move the x axis nema and ballscrew outside of the gantry "legs" so if needed I can support the beams the x axis supported rails are sitting on.

[JAZZCNC]

Will probably move the x axis nema and ballscrew outside of the gantry "legs" so if needed I can support the beams the x axis supported rails are sitting on.

Much much better and far stronger regards the frame and gantry. Yes bring the ballscrews outside the profile has this will make it stronger still not only allowing supports but brings the screws closer to the gantry and X axis bearings.
I would re think the Z axis as that's very limiting how you have it designed now.? Look around the site and you'll see much better setups.

[Robin Hewitt]

I know you want to get value for money on the Z axis screws and slides, but there really is no point lifting the head by much more than the distance from gantry to table. I suggest you could shorten the stroke and make everything neat and tidy.

[Washout]

Latest pic:

Moved the rails on the Y axis to the top and bottom of the gantry (thanks Mocha) and I assume that was what you also meant Jazz.

Thanks Robert - I have my eye on a set of ballscrews that are 1050mm, 550mm and 350mm currently (machined), so the frames/axis are kind of sized to match. I might rethink that though when I have the main structure layout set.

This is all extrusion currently, but if its sufficient to cut plate then I'll make some replacements for various bits once its up and running. I'm also a bit concerned about the rigidity of the extrusion couplings (I guess I could add some structural epoxy to help with that also, as it seems to work for Lotus, Jaguar and Aston Martin to name a few ;).

[Jonathan]

Moved the rails on the Y axis to the top and bottom of the gantry (thanks Mocha) and I assume that was what you also meant Jazz.

Jazz will have meant putting the Z-axis rails on the plate that moves up and down, not the bearing blocks. Like mine (old picture):

64476448

However you do it, as Robin said, with the short distance between the chuck and the bed there is no need to have so much travel. You're much better off spacing the bearing blocks out further on the linear rails (this sacrificing the travel) since this reduces the magnitude of the force on each, leading to higher stiffness.


Maybe a silly question, but is the tool chatter etc linked to harmonic resonance in the frame essentially and if so would assymetric mass damping help i.e. attaching weights to the extrusions in irregular places to stop it singing like a tuning fork?

Yes, that or avoiding symmetry in the frame can help to prevent resonance issues.


I have my eye on a set of ballscrews that are 1050mm, 550mm and 350mm currently (machined), so the frames/axis are kind of sized to match.

Bear in mind Chai (linearmotionbearings2008) on eBay will cut and machine them to any length you want and the price is just based on the length so there's no need to stick to a particular set of sizes just because that's what he's happened to list on eBay for someone else.

[JAZZCNC]

Latest pic:

Moved the rails on the Y axis to the top and bottom of the gantry (thanks Mocha) and I assume that was what you also meant Jazz.

No not at all like that.! . . . I was talking about the X axis ball-screws and a completely different Z axis design not rail placement.
Not got time to explain now but will do later if you like or someone doesn't beat me too it.!

[Washout]

Ah thanks Jonathan - that makes sense and gives a lot more clearance under the gantry. Questions though:

I thought having a narrow Y axis rail spacing wasn't desirable due to the leverage that can be applied when under x axis cutting motion? Sorry - I see you already addressed this in your reply....

Also, how do you mount the setpper motor on that design - between the rails on the back of the plate?

@ Jazz - no worries, forums aren't the best medium for communicating or understanding design concepts - shame there isn't a fag packet app built into vbulletin ;) If/when you have time to explain I'd be more than grateful for a longer explanation and I'm happy enough to work in the 3d virtual world (it's cheaper) whilst I learn the concepts.

Given that I've hoovered up many supplier links from various threads on this forum, this build is coming in well under my initial budget, which means I think Option 3 (DIY) from my original post is now looking like my decision thanks, to you guys.

Oh and is there a place for discussing 3D Printing on this forum I haven't seen yet?

Cheers


Chris

[Washout]

Here's the latest iteration - changes are:

- Re-structured the Z axis to kind of match what Jonathan showed in his photos and some general tidying up in that area.

- Made the corner posts on the bed taller and moved the x axis ballscrew assembly to the same height as the rails and outboard (did the same on the other side as well to keep a dual ballscrew option open).

Still to do - add in the bearing pillow blocks, which will probably mean some changes to some areas to accomodate and anything else I am missing, once you guys have finished ripping into it :wink:

[Washout]

Hi Gents,

Paying job has been keeping me busy, but I have made some more alterations to the design to get to the state in the diagram attached. Actually I'm going to go with double thickness (60mm) for the four "corner posts" on the bed/frame and also thicken the T-Slot bed to 30mm even though the bed will be supported in 4 places.

Still not sure about the configuration of the Z Axis, as I see using a belt/pulley arrangement with linear rails and the stepper mounted to the side opposite to the spindle seems to be the accepted "strong" option. Also ignore the spacing of the bearing blocks as that will change or be replaced if I go with the arrangem,ent in the last sentence.

Questions (there seems to be more and more):

1. For the X Axis and from other posts I am guessing that when ordering ballscrews and assuming that I will be mounting the bearing/support blocks on the outside of the "corner posts", I need to order a 1605 ballscrew that is 1081mm with standard machined ends (total length of frame is 1,000mm currently) i.e. for 1605 its length of screw thread plus 81mm (30mm+39mm at the fixed end and 12mm at the supported end).

2. The BK blocks appears to be an FK mounted into the BK block housing, so if I order Chai's standard 1605 package, then will I be able to take the FK out of the BK block and use it in say an MBA-12C motor mount from somewhere like Zapp?

Cheers



Chris