PDA

View Full Version : BUILD LOG: 3-Axis CNC Mill Project



Hobgoblin
25-08-2012, 06:29 PM
3-Axis CNC Mill Project

After spending a few weeks contemplating ideas for a new hobby project; then longer deciding on a design I was happy with I finally got my hands on the Aluminium extrusion profile wanted to build the frame.

Aims are to have a complete unit, machine base and storage, 3 axis CNC control with Mach3 software, A3 paper size capability, High precision with zero backlash, hardened linear rails ball screws and stepper motors and finally a special purpose milling spindle.

First making a sketch of the idea:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0124e.jpg

Then realising things look a lot smaller on paper when I get the aluminium home:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0102.jpg

Anyway, first up is to end tap the profiles required M8, thankfully its extruded with the 6.8mm hole required for such:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0104.jpg

That done:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0109.jpg

It was on with marking out the holes to assemble the frame:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0111.jpg

And then on with drilling them:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0112.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0113.jpg

Hobgoblin
25-08-2012, 07:06 PM
Then on to a mock assembly to make sure it looks ok:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0114.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0116.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0118.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0119.jpg

Perfect, a solid accurate level frame with enough room underneath the table for housing the control, pc, drives, tooling etc. Finally I laid the aluminium extrusion profile I’m using for the machine table on top:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0120.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0121.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0122.jpg
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0123.jpg

Next up is to finalise the design for the hardened linear rails, linear bearing guide blocks, ball screws and stepper motors; that said I can then cut this profile to suit, and make the carriage for the x axis as shown atop the sketch:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0125.jpg

Sit down time with a good bottle of ale:
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0127.jpg

Hopefully when its complete I’ll be able to mill 2D images to 3D, and produce high precision parts.

Thanks.

boldford
25-08-2012, 11:22 PM
Wouldn't the frame be stiffer with a few diagonal members added?

Hobgoblin
26-08-2012, 11:24 PM
Wouldn't the frame be stiffer with a few diagonal members added?


No, you must be unfamiliar with my materials, the stuff im using is stiff and strong enough for what i need it for.

//EDIT - (Two months later to the day) Oh how wrong and misguided i was, keep reading and you'll see progress from this day forth hehe.

JAZZCNC
26-08-2012, 11:53 PM
No, you must be unfamiliar with my materials, the stuff im using is stiff and strong enough for what i need it for.

Think you'll find he's correct and Yes I am familiar with your materials.!!

What your not accounting for is the Mass of the gantry and it's inertia.?
It doesn't so much matter what your cutting the Mass of the gantry will have a big affect and actually if soft material then usually means high feed rates which affectively makes the gantry a very heavy thing indeed which will put more stress upon the frame trying to twist it and push it out of shape.!! . . . At this height and narrow width and also from the look of your doodle showing the high gantry sides I wouldn't be one bit surprised if it didn't try to tip it over when rapiding at even mild speeds.? . . . I would certainly think about bolting it down or have plenty of ballast low down.!!

For the little extra cost it would be well worth it to brace the frame now rather than find out it's twisting and moving around while cutting then have to do it later and go thru all the shimming and setting up again.!!

Jonathan
26-08-2012, 11:54 PM
High precision with zero backlash

What sort of accuracy are you expecting and how do you propose to get zero backlash, or is that meant to read 'near enough zero'?

How wide is the machine, about 600mm between the bearings? It appears from your initial sketch that you are only using one ballscrew to move the gantry which I'm confident you will regret since you describe this as a CNC Mill, not router, implying you wish to use it to cut metals.


Wouldn't the frame be stiffer with a few diagonal members added?


Yes, it would be significantly stronger. If the frame is stiff enough without (which it's not) then that just means smaller extrusion could have been used with diagonal pieces for the same strength but lower cost. Still, they can be added at a later date.

Web Goblin
27-08-2012, 09:54 AM
I agree with the guys on this one. That frame is going to move. As Jazz says given the height of the frame this will probably need fixing down.

Hobgoblin
27-08-2012, 11:05 AM
Just to clear a few things up.


I do welcome the feedback and criticism good or bad as it helps to keep in mind the planning and checking process.


The purpose of this project for me is, first of all just to be able to, and then for as said engraving, hard woods, plastics (acrylic lexan etc) and possibly aluminium. This isn’t unachievable with this setup, however trying to skim a cylinder head or reproducing engine parts is.


Achieving a level table, parallel axis and zero backlash can be achieved with the right setting equipment, depth micrometers height vernier's DTI's (which I can borrow freely as I do this stuff for a living but on a much greater scale), as for backlash if any is present when finished this can be compensated for in the software I intent to use to control the machine.


Vibration and tool chatter will be a problem in certain situations and of course it won’t be as easy to overcome as removing some of the inserts from a cutter to stop such resonating because this isn’t a large machine with a large face milling cutter.


But I can only try to design what I can afford for the intended purpose and strengthen it where possible. Reducing depth of cut, feed per tooth per rev, spindle RPM tweaking all the basic parameters if in a situation where I’m still getting chatter/vibration.


I was actually quite impressed by the overall mass of just the base once it was assembled and it certainly isn’t flimsy stuff even the 20x80 is very strong.


I have considered making some brackets coming off the bottom corners of the frame that, once the machine is levelled at the base, will also bolt into the ground.


In the final stages of designing the next steps at the moment, all the planning for the traversing carriage, the guide rails, ball screws, stepper motors, and brackets needs to be carefully considered so that I don’t encounter any parts fouling on others.


The machining area should be about 400x500x150mm when complete but thus depends on how things develop; there will defiantly be room for A3 though.


Thanks.

John S
27-08-2012, 11:57 AM
as for backlash if any is present when finished this can be compensated for in the software I intent to use to control the machine.




Thanks.

Backlash compensation in software is a fudge at best.
When a cutter is doing a circle, at the 4 quadrant points the tool is free to move and will grab in the climb milling direction whilst the software is working out which way to go.

JAZZCNC
27-08-2012, 12:06 PM
Just to clear a few things up.

I do welcome the feedback and criticism good or bad as it helps to keep in mind the planning and checking process.


Thats great because I'm sure were going to give get plenty before your done. . :joker:

I hear what your saying regards usage but still that Gantry has Mass which will try to bend and twist the frame. Even when engraving or V-carving etc the small and sharp direction changes will rock,twist and resonate the frame which all transfer back to the job.! . . . Extra bracing will help with this is what where saying.!!

Re-Backlash then it ain't easy getting true zero and software comp isn't a very good answer or backup IMO but that said the Chinese ballscrews will be more than good enough for this machine and it would be pointless going to all the trouble of trying to achieve true zero has it requires a much stronger base frame than what this design provides.


Keep up the good work and pics coming.!

Robin Hewitt
27-08-2012, 01:01 PM
You could turn the open frame into a cupboard to keep your knick-knacks and controller in. If you cut Melamine faced board to a good fit on 3 sides it would make that base totally rigid.

Beware the temptation to make the gantry tall for that large lump you think you might want to cut some nebulous time in the future. Height costs you rigidity.

Your tool is unlikely to extend much more than an inch from the collet so you'd only be scratching the top of that lump anyway.

Hobgoblin
27-08-2012, 05:26 PM
I've made a little more progress this afternoon.


I've finished a mock assembly of the carriage. This is subject to being shortened on the height as much as the finished design will allow so that I can still get a good travel in the Z axis (up and down).


http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0129.jpg


http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0130.jpg


As a contingency factor I’m considering the prospect of using 4 linear rails in the X axis rather than just 2. This isn’t my first choice though as it costs more. BUT if I can get good stability without drastic effects from inertia of the carriage rapid traversing then I'll leave it with just 2. There will always be the option to upgrade to the 4 though so that will be a bonus if like said required.


Just doodled over a picture if any can can’t understand what I’m babbling on about, the square blocks are the linear bearing blocks and the “?” are the extra ones factored.


http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0132a.jpg


http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/IMAG0133.jpg


The side supports like said are subject to being shortened as much as I can with the top brace to follow suit so the set up is as pictured, just shorter.


Next thing to do will be to order some hardened rails, support brackets and linear bearings and take things from there as and when funds allow or the missus signs my pass to take over the kitchen.

JAZZCNC
27-08-2012, 05:53 PM
Ok I'll shoot first.!!

Please take what I'm about to say has help rather than pulling a design down.!!

That gantry is miles too narrow and it 99.9% will twist and rack even if you use 4 bearings in the arrangement you show. It will 100% rack if you use a single down the centre ballscrew using this arrangement.?
It does need 4 bearings but not like you have them drawn. They need to be in-line side by side creating a wide foot print.

The gantry cross bracing (Y axis) is no where near enough and it will flex with Y axis sides ways movement, 2 of those pieces would work much better.

If you are going to use this narrow profile then I would recommend you fasten wide plates at the bottom for the bearings to mount side by side.
Also some corner bracing wouldn't hurt.

Don't know where your inspiration for this machine comes from but it has a very Merchant dice feel to it and if so then let me tell you they are rubbish design which flex and move around like quivers jelly.!!

Hobgoblin
27-08-2012, 06:07 PM
You could turn the open frame into a cupboard to keep your knick-knacks and controller in. If you cut Melamine faced board to a good fit on 3 sides it would make that base totally rigid.

Beware the temptation to make the gantry tall for that large lump you think you might want to cut some nebulous time in the future. Height costs you rigidity.

Your tool is unlikely to extend much more than an inch from the collet so you'd only be scratching the top of that lump anyway.

Very valuable information in all three paragraphs, your first, I had planned on doing, housing the pc tower, drives, tooling etc.

Thanks Robin.:thumsup:

John S
27-08-2012, 06:43 PM
it has a very Merchant dice feel to it and if so then let me tell you they are rubbish design which flex and move around like quivers jelly.!!

It's Chivers jelly you cretin, it quivers BEFORE you put the supporting custard on it.

JAZZCNC
27-08-2012, 06:51 PM
It's Chivers jelly you cretin, it quivers BEFORE you put the supporting custard on it.

He he.!! . . Who gives a fooook it still going to wobble like my tits on my old triumph T120.!!. . . . Any way we couldn't afford custard had to make do with Carnation milk.!!

Hobgoblin
27-08-2012, 06:51 PM
Ok I'll shoot first.!!

Please take what I'm about to say has help rather than pulling a design down.!!

That gantry is miles too narrow and it 99.9% will twist and rack even if you use 4 bearings in the arrangement you show. It will 100% rack if you use a single down the centre ballscrew using this arrangement.?
It does need 4 bearings but not like you have them drawn. They need to be in-line side by side creating a wide foot print.

The gantry cross bracing (Y axis) is no where near enough and it will flex with Y axis sides ways movement, 2 of those pieces would work much better.

If you are going to use this narrow profile then I would recommend you fasten wide plates at the bottom for the bearings to mount side by side.
Also some corner bracing wouldn't hurt.

Don't know where your inspiration for this machine comes from but it has a very Merchant dice feel to it and if so then let me tell you they are rubbish design which flex and move around like quivers jelly.!!

I fully take on board your comments jazz.

I can see now with slightly less blinkered eyes that the opinions so far have a good grounding behind them.

My only option is to start from the original sketch, assess and improve where ever i can.

Even if it serves no greater purpose than an expensive learning curve and thats the only benefit ive gained from doing this project at least i can put that expirience under my belt.

As for the merchant dice thing yes i did look at those, and designs from youtubers, neo7cnc and hamrx8 for some of the inspiration for my design, that and i was able to come by the aluminium i have used for free has left me with what i have.

Im not too disheartened, but will glady accept any furth help your willing to offer.


as for the greater foot print for the base in x, how about if i mount two pieces of 40 80, long side vertical, inside the frame, top edge of the 40 on the same plane as the under edge of the table top, and on these pieces mount each side a linear rail to meet 4 linear bearings on the carraige that way?

thanks again.

JAZZCNC
27-08-2012, 07:32 PM
The base is not so much an issue and easily sorted with bracing or like Robin suggested stiffening with panels.

The tall narrow gantry is the biggest weak area I see and is where all the action takes place. It's also where all the other machines of similar design fall down.!

If your having a single central screw then you need a wide bearing footprint to help counter racking when cutting at the outer edges. The wider the better but obviously this loses travel so it's always a compromise.!

The other area that is often over looked and is THE most important part of the machine IMO the Z axis. It does all the work so if this is flimsy then it don't matter if the rest of machine is carved from one lump of granite the finished result will be shite.! . . . Obviously this is fastened to the Y axis and again if the gantry is flexing around because it's too tall or narrow and under braced then so will every thing attached to it.!. . . . It's also what Robin was pointing out.!

It's really just a case of make the X axis bearing foot print has wide as you feel you can spare and make the gantry only has high as you really need. If you need high then build very rigid with loads of bracing.

Don't under estimate the inertial forces of the gantry and Y/Z axis they can be quite high and will easily rock and flex a high narrow gantry at half decent cutting and rapid speeds.
This is what plagues the Merchant dice type offerings because the motors and Ball-screws are capable of so much more than the frame will allow and the finished cut is drasticly reduced if run too high.!! . . . So therefore the machines run feed rates well below there components potential to give anything like a decent finish and stop the frame from falling apart.!! . . . . This soon becomes very limiting and restrictive and why most sell them shortly after buying.!

My suggestion is you Stand back a little have a think and look about at other machines and throw a drawing or two our way for checking over. Has you've probably seen already where all more than willing to help and advise with honest opinions.

Hobgoblin
29-08-2012, 09:49 PM
Starting to see the error of my ways so far. Taking time out to work on some things. Will update once progress is made.

Hobgoblin
31-08-2012, 04:43 PM
http://img.photobucket.com/albums/v248/ROBOTIX/3-Axis%20CNC%20Mill%20Project/gantrtysupportidea.png

Does this look like a good place to start for improving the gantry, im thinking of shortening it to 450mm from 600mm, adding these supports shown, and also large braces at the top in Y, and corners all round.

thanks

Hobgoblin
06-09-2012, 10:16 PM
http://videobam.com/FeuSv <-- link to video

Ive knocked this up in solid works, there is now 200 gantry plates, and the gantry profile is now 425 long not 600.

Any advice, input welcome and required lol thanks

Hobgoblin
09-09-2012, 09:29 PM
Big thanks for Jazzcnc's input in my project, I'll be referencing his ideas as i get further into the project, so far the guy has gone far far out of his way to lead me on the right path.

Hobgoblin
24-09-2012, 07:12 PM
This far now, this was supposed to be a re draw. But I think ill be going for a re re draw just to be sure all my parts right.

Thanks

Hobgoblin
24-09-2012, 07:26 PM
The eagle eyed among us will of spotted my ball nuts missing for x, thats because im shit at cad and still learning and they have disapeared, so i put them back in.

Thanks.6985

Hobgoblin
17-10-2012, 09:51 PM
If I am driving my X axis with two 16mm diameter x 10mm pitch screws via a belt and pulleys, do i use a 1:1 ratio? so the pullys on the screws and motor each have for example, 40 teeth?

Jonathan
17-10-2012, 10:09 PM
Yes, for the vast majority of things 1:1 is the best choice to get good speed and acceleration with 10mm pitch screws. There may be occasions when you need to use 1:2 (e.g. motor 15:screw 10), but it would need to be something very detailed to warrant that. Also there's no need to use such a large pulley, it will only increase the inertia of the system and reduce your acceleration as a result. 30T is a good choice since then you can use 1:2, in addition to 1:1, without having to use very small pulley.

Hobgoblin
17-10-2012, 10:15 PM
See now i just feel really thick. I had in mind 5mm pitch, 15mm belt. Ive already been having a lift with this from jazz and it has helped me to be in the right direction but i still dont fully understand this aspect of the machine im building. Not that im ignoring his advice i just thought id throw the Q on here and see what comes of it. Your last sentence went straight over my head jonathan, sorry :cower:.

Jonathan
17-10-2012, 10:31 PM
If you use a 30T pulley on the ballscrew then you can put a 30T pulley on the motor to get 1:1, or use a 15T pulley on the motor (which is unlikely to be needed much if at all, but it's good to have the option) to get 1:2 ratio. It's best not to use smaller than a 15T pulley since the belt will wear quicker and you may not have enough teetth engaged with the pulley if it is too small.

The moment of inertia of a cylinder (which is essentially what the pulley is) is proportional to its squared, so a small increase in the pulley diameter will cause a relatively big increase of the inertia. The torque required from the motor is proportional to the inertia, so you want to minimise it if possible - hence use the smaller pulley, 30T instead of 40T.

Hobgoblin
17-10-2012, 10:47 PM
This has helped some what. Aim for 1:1 but leaving the option for 1:2, got it :encouragement:.

I'm up to working with Chai to finalise a parts list at the moment, Then its save up time. Hopefully I can get my china bits before xmas.

Hobgoblin
26-10-2012, 09:12 PM
Before xmas for my order of bits now looks unrealistic. Other more important things have had my pocket money, new helmet for the bike being the biggest.

Anyway, I have now got all the extrusion I need. I was 1 piece of 20x80 x 600 missing for my new design to work. And My side, vertical, pieces of 20x80 were 600 high, have now been re cut down to 425 high.

I have a fair bit to get on with though.

Photos to come.

wilfy
28-10-2012, 12:34 AM
looking forward to seeing this finnished.. it looks like your going to get this done before i finish mine and as i've just learnt your about 15 minutes away from me i may have to send some work your way :D

Lee Roberts
28-10-2012, 01:11 AM
as i've just learnt your about 15 minutes away from me

Fello Wigan'er here also!

wilfy
23-12-2012, 02:05 AM
how you getting on with this fella?

Hobgoblin
23-12-2012, 11:31 AM
On hold at the moment, currently on with the purchase of a new motorcycle so thats taking up most of my time right now.

wilfy
23-12-2012, 11:47 AM
ah well you'll have to pop down to mine soon then as i've got all my china goodies here already