CNC routing

[stef]

I use EMC2 (Linux) , and CAMBAM to get the CAM files generated.
But then again, i am not set on any particular software, as long as it's free.
Ballscrews are 25mm diameter, 10mm pitch. 2 of them on the xaxis, 1 (1100mm) on y and 1 (400mm) on z .

 

[RussianRouter]

Paul,its basically a swindle,added it to cart and it started asking for motors added and the 2nd pair are expenive.

 

[stef]

if i may offer a little advice: dont rush your purchases.
no need to buy everything from the start (motors, screws, drivers, psu etc...)
motors are a standard format, nema 23 or 34 usually, so you can build the machine without buying anything too expensive.
or wait for the right deal on ebay.
i managed to get my hands on some gecko G202s, for around £50 a piece. They are far better than the leadshine on offer from zapp.
i got them last week.. and you've seen how far i am in the build !

it's tempting to fork out for good screws, good motors, good drivers, and then
have them sit in a garage for 2 years, before decided to sell them at a loss !
start designing, then building. the buying is the easy bit, you can leave it until the end.

well, that's my advice anyway.

 

[RussianRouter]

What do you reckon on this,Stef

http://cgi.ebay.co.uk/3-Axis-CNC-Router ... d4fefabc6d

I'm not after a super machine but one that is capable of handling a small bed.

 

[stef]

looks cheap enough.
but as i said, those are plentyfull.
i'd first consider what type of design you want (moving bed or gantry based)
then start making it.
the screws and linear bearings/guide, are far more crucial than the motors and drivers.
but then it's also a good idea to keep final costs in mind. but dont rush in buying the last pieces of the puzzle before you've even laid out the first.
imagine, you come across dirt cheap screws and rails in the next month, but turns out your motors are just a bit on the small side...
if you wanted to do this by the book, you'd have to workout the loading on each motor, before actually buying it. and this loading will depend on the screw, the type of design, the type of rail, etc...
plugging the motors to the controller and get them stepping took me 2 hours. building a small machine is likely to take you about 2 months.
i'd also consider quality of the rest of the equipment. if it's for a poorly built MDF machine with M8 threads as screws and ball bearing guides, then yep, those drivers are of equal quality.
but if you intend to invest in decent screws and guides, forget about those drivers. get some Geckodrives, or at a minimum, some leadshines.
I built the drivers myself on my first machine (based on L297s), and i can tell you that they can make or break a machine. for the big machine, the one on the picture , i was prepared to pay over £600 for the drivers, that's what i budgetted for. then i got a bargain of ebay !

 

[RussianRouter]

http://www.youtube.com/watch?v=MLTAeE6j ... re=related

 

[stef]

a machine like this looks extremely accurate.
may be small, but probably decent screws and rails.

I am hoping for this sort of precision on mine. hope i will achieve it !

 

[RussianRouter]

Stef if you don't mind me asking what is the work it will be undetaking?

I think a flatbed and an overhead gantry adds to its versatility.

 

[stef]

Stef if you don't mind me asking what is the work it will be undetaking?

I think a flatbed and an overhead gantry adds to its versatility.

It will be mainly for wood.
but it should be capable of aluminium and possibly steel.
the configuration depends on the size, mainly. above a given size, a moving gantry makes more sense than a moving bed.

 

[wobblycogs]

I hope your going to post loads of pictures RR - really looking forward to seeing how this project pans out.

I've been planning to build a CNC for years, as soon as I get the bathroom I'm working on finished and the router table is built I think I'll probably start. I've always said that I'd build a gantry type design. The gantry design feels more complex to get right but the smaller foot print is a real bonus.

 

[chipchaser]

Hi George,

I don't know whether you have found this site yet, the link takes you to the Fora

http://www.cnczone.com/forums/index.php

It may not be the best site in the world, although I believe you can find almost everything you want to know there and it must be the biggest - what else would you expect from the USofA

You can see many types of machines, lots of discussion of the pros and cons of alternative machine layout and there are various plans, some free. I mainly visit the Woodworking Machines Forum.

If you are thinking of self building it may also be worth looking at

http://www.solsylva.com/

Even if you don't buy his plans there is a lot of good info there. Oh, and I like this one too especially the chain drive for a long movement axis.

http://buildyourcnc.com/default.aspx

HTH and apologies if you've been there before.

Graham

 

[stef]

that's my build log, on cnczone.
http://www.cnczone.com/forums/showthread.php?t=101685

 

[RussianRouter]

I've always said that I'd build a gantry type design. The gantry design feels more complex to get right but the smaller foot print is a real bonus.

With a lot of searching and looking the gantry is essential to the lift of the cutter away from the job,I've seen partial builds on you tube that are going to fail at the first hurdle because they have not implemented the Y axis.

Its pretty straight forward and there's a couple of ways in doing the gantry slide,however as mine won't be as large as Stef's the princple is the same on a smaller scale...just less movement of the cutter in the X and Z axis depending on the width length.of the bed.

The width of the bed on this bed will be 24" about 30 ?" I'll decide on the build when it starts.

I'm trying to keep the cost down to a minimum without losing stability and accuracy.

 

[RussianRouter]

Hi George,

Thanks Graham.

Been on more sites than you can imagine over the last couple of days searching for design inspiration,last night I must have spent 2 hours on youtube browsing and watching the homebrewers machines.

There are some real idiots out there,a guy was using his hand power drill on his effort as the cutter. :lol:

 

[stef]

I've always said that I'd build a gantry type design. The gantry design feels more complex to get right but the smaller foot print is a real bonus.

With a lot of searching and looking the gantry is essential to the lift of the cutter away from the job,I've seen partial builds on you tube that are going to fail at the first hurdle because they have not implemented the Y axis.

Its pretty straight forward and there's a couple of ways in doing the gantry slide,however as mine won't be as large as Stef's the princple is the same on a smaller scale...just less movement of the cutter in the X and Z axis depending on the width length.of the bed.

The width of the bed on this bed will be 24" about 30 ?" I'll decide on the build when it starts.

I'm trying to keep the cost down to a minimum without losing stability and accuracy.

sorry to break it to you so bluntly, but it's not that straight forward.
eventhough i conceide that going smaller than mine will spare you a lot of hassle.
one of the main thing you will be fighting is flexing of your support. leveling and squaring is also another big issue. there are flex calculators on the web that will help you choose the right material/profile for the frame/structure. but you just cant pop into b&q and pick up any piece of material. not if you want accuracy.

If it's a hobby machine, only to be used in a dry environment, then MDF is fine. as long as you build in torsion boxes and a clever frame.
if it's a advanced hobby, and in a garage, then it's steel or aluminium. even with steel, balancing things out is a little tricky.
you will need to "machine" or flatten all the beams onto which guides are resting.
dont expect to spend less than £1k on a 24x24 machine. that's if you want to machine aluminium.

 

[RussianRouter]

Hi Stef

When I say accuracy I mean within a tolerance factor,yours will be precision becaue ofthe nature of the work you want.

I'm building this fo one reaon and that is to cut out shapes and carvings to affix to wood or inlay in wood,it is not being desinged for doing dovtail or any other joint but I'll see how it fairs with a couple when I get my head round operating the software.

 

[stef]

Hi Stef

When I say accuracy I mean within a tolerance factor,yours will be precision becaue ofthe nature of the work you want.

I'm building this fo one reaon and that is to cut out shapes and carvings to affix to wood or inlay in wood,it is not being desinged for doing dovtail or any other joint but I'll see how it fairs with a couple when I get my head round operating the software.

fair enough.
you still have to be carefull though. my first machine (salvaged most parts, ball bearings as guides, M8 as leadscrews) still cost me over £300. I had to give up tweaking it, because it was just not right from the start. it would move the router fine, but when you cutout a shape, you never do it in one pass. the software works out for you that you need overlaying pass, say 5 times 2mm for a 10mm deep cutout. if your machine is not spot on, if it misses steps, or wobbles a bit, then the 5 overlaying pass are not on top of each other, and the piece is messed up.mine messed up so many times, it would take me 5 pieces to get one right.

It really isnt a job to be taken lightly, and if you really want/need results, then spend on good quality components. if you just want to have a go and see what you can do on a budget, then it's a different game.

 

[RussianRouter]

Ok lets see if I've got this right in my mind?

Duh! didn't think that one through,sorry

I'll have a rest. :lol:

 

[stef]

Ok lets see if I've got this right in my mind?

Lets supposing I'm making a box 12"x8",I want a dark oval shape inlay in the lid of the box.

I take two pieces of wood exactly the same size,one of those pieces is the lid of said box the otherr piece willbe the dark ovalhape.

I place them on top of each other and affix to the bed,set the cutter in motion,job finished.

Are you saying that those two pieces of wood will not be identical ie shape&cutout after its made about four or five pasess?

I am not sure if i follow your plan to inlay the wood.. did you take the diameter of the bit into account ?
re: the number of passes,it will depend on the thickness of the wood, and how hard it is.
Think about routing like you would handheld: your machine may not even be as strong as you holding the router, so you have to go easy on it: to cut through a 10mm plank you will probably want to make 5 passes, 2mm each time. each pass is on top of the previous one, in theory. except that to achieve this, your machine has to be pretty precise (repeatability). it's easy for a stepper motor machine to be loosing steps, either from the electronics, or lack of power, or some mechanical gremlin, etc..backlash is probably the most obvious reason why the two passes wont be on top of each other. which is why you generally have 0 backlash nuts on these machines..

overall, designing a cnc is to make massive amounts of compromises. even if cost isnt an issue, you will still have many compromises which will get sorted depending on your specific application.
you can have tall gantry, small legged gantrys, twin screws, single screws, steppers, servos...

 

[chipchaser]

Hi Stef,

That looks very good, impressed by the steel frame. What accuracy are you aiming for with this one?

What did you make your first one from and how accurate was that?

Have you found EMC ok?

Look forward to seeing more of your machine as it progresses

Graham