2 Attachment(s)
From Aylesbury, and looking for some help
Hi
I make kitchens in a very non-cnc kind of way, and I want to use CNC techniques on a simple woodworking machine. Unfortunately, although I have a reasonable grasp of how CNC works, and I understand (mostly) what the various parts are, I have no practical experience at all and so I am here hoping someone can help me or at least point me in the right direction.
Let me tell you what I want to do- I want to be able to move a fence (in the woodworking sense) of about a metre long, over the surface of a table so that the fence remains parallel to a fixed rail. Illustrations below
Attachment 4256
Attachment 4255
I would like to be able to position the fence to an accuracy of 0.1mm, if at all possible. A total travel of about 800mm would be enough. Once in position I need it to be able to resist the force needed to push a piece of board material against it so that it can be cut by a saw moving along the fixed rail (which I will do by hand).
I would also like to be able to make a trial cut, measure the resulting piece, then tell the computer (I have a machine that runs XP at the workshop) what the length of the piece is, then tell the computer what length piece I require, and have it move the fence accordingly, so I will need some software too.
I really don't know what the best way to achieve all this is, so am hoping someone here can help, either on this thread, or by email, or in person (an expenses budget is available
Re: From Aylesbury, and looking for some help
Sounds like you need a CNC Arbor/ Swing saw, then no hand cutting at all. Or just buy a swing saw set up on a roller bed? I would keep this one simple myself http://www.mycncuk.com/forums/image/...AASUVORK5CYII=
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
2e0poz
Thanks for that, but the cutting aspect is absolutely no problem at all. It's positioning the movable fence (green element in the sketches) that I need to get sorted out. I'm happy to keep it simple, as long as it is effective. The picture (if that's what it is) in your reply doesn't work, BTW.
Re: From Aylesbury, and looking for some help
You don't say how long the fence is but you will need either two support rails for it and one feedscrew, ballscrew actually if you want to keep to 0.1mm accuracy, or better still two ballscrews linked by timing belt to stop the fence crabbing.
Add to this a stepper motor, driver and power supply.
Your XP machine could run a trial version of Mach3 in that you only need MDI input and do not need to run any lines of code. You could even use a custom screen.
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
John S
You don't say how long the fence is but you will need either two support rails for it and one feedscrew, ballscrew actually if you want to keep to 0.1mm accuracy
Surely a trapezoidal screw will have sufficient accuracy. You can get them in similar accuracy grades as ballscrews. C7 grade would be fine. Since speed probably isn't an issue the efficiency of the ballscrew is not an advantage and with this set-up it is very easy to compensate for the greater backlash. Just seems a bit over the top to use a ballscrew to me...
Re: From Aylesbury, and looking for some help
C7 is crap.
C5 is the bare minimum I'd use and from some of the offering I have seen lately with nuts that rattle up and down, my sissor jack on the truck is better made <g>
The OP wanted 0.1mm tolerance that's 4 thou and given his sketch with a 800 mm travel it looks like the fence is around 1M, so with a screw either end to keep this tolerance you have a max of plus or minus 2 thou to play with over 1M or 39" [ excuse mixed measurements ]
If these tolerances are 'needed' as opposed to 'I'd like' then you have to poppy up or someone is chasing errors when it's not needed.
Re: From Aylesbury, and looking for some help
The fence will indeed need to be around 1 metre,( sorry, should have said). The 0.1mm accuracy would, while not essential, certainly be worth paying extra for, 0.5mm would be disappointing and I doubt it would be worthwhile building it.
Re: From Aylesbury, and looking for some help
The difference in accuracy between C7 and C5 is not that much. Over an 800mm length you're surely going to have more inaccuracy caused by deflection of the frame/tool and the resolution / step accuracy of the stepper motors to an extent. The piece that the OP said is about 1m long will need to be strong and rigidly supported at the ends to not deflect by +-2thou.
http://www.mycncuk.com/forums/showth...curacy-V-Price
The C7 grade screws I have all seem fine, though clearly I have not got the kit to measure them that accurately. Either way leadscrews with the equivalent to C5 should be cheaper than ballscrews.
I would definitely use a screw on both sides linked by a timing belt. You could use two motors but then it's more difficult to keep it square. A small amount of skew will stop you getting 0.1mm tolerance. Why do you want that tolerance for kitchen worktop? How are you going to measure upto 800mm long to the nearest 0.1mm?
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
Jonathan
I would definitely use a screw on both sides linked by a timing belt. You could use two motors but then it's more difficult to keep it square. A small amount of skew will stop you getting 0.1mm tolerance. Why do you want that tolerance for kitchen worktop? How are you going to measure upto 800mm long to the nearest 0.1mm?
It's not for worktop, it's for cabinet components. As for the accuracy issue, let me explain it thus- if I make a cabinet with sides 720mm high by 565mm deep, and a base which is 564mm wide and 510mm deep (the difference in depth is for the back panel which is rebabated into the sides at 510mm from the front edge, but which is fixed to the back edge of the base panel) and the sides are actually 565.5mm deep and the base 509.5mm then there will be a 1mm discrepancy where the side meets the base at the front. That much discrepancy would be clearly visible, and even more obvious to the touch. Also the cabinet will be slightly out of square if the top of the side is 565mm and the bottom of the side is 565.5mm
At the moment when such discepancies occur I have to sand the proud part down until it is level with the other part (my cabinets are made from birch ply rather than laminate, so this is feasible) but I would much prefer not to have to do that and am prepared to go to some trouble and expense (I'm expecting this to cost me a good few hundred at least) to end up with an accurate machine.
Re: From Aylesbury, and looking for some help
It looks like repeatability is what you need, not accuracy which is good as even a poor quality screw will still be very repeatable.
Maybe make it so that you can easily put the base panel back on the machine to take a tiny bit off untill it fits?
Surely you would plane the proud part down, not sand it. Shouldn't take long to plane off a mm or so.
Have you considered making a complete CNC router ... then you can make it cut the holes rebates etc automatically.
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
Jonathan
Maybe make it so that you can easily put the base panel back on the machine to take a tiny bit off untill it fits?
It depends on whether it's still parallel- the accuracy is not just for ease of fitting the parts together, it's also needed to sure that the opposite edges are parallel, so that the compononent has all four corners at right angles.
Quote:
Originally Posted by
Jonathan
Surely you would plane the proud part down, not sand it. Shouldn't take long to plane off a mm or so.
A decent random orbit sander is faster, and the discs can be changed quickly when they wear out. Planer blades would ned to be TCT as the material can be quite abrasive
Quote:
Originally Posted by
Jonathan
Have you considered making a complete CNC router ... then you can make it cut the holes rebates etc automatically.
I have certainly thought about it, even been tempted once or twice. Holding the material down would be a problem, birch ply does not lie as flat as most kitchen cabinet materials, and a powerful vacuum pumo and system would be needed. There's also the issue of needed to learn quite a bit of software, whereas at the moment I can keep most of the numbers I need in my head, and the 8'x4' sheets to cut down in a quite logical sequence, and if one of the boards has a nasty knot in it I can easily cut around it to avoid it.
One fo the nice things about a machine such as I described in the original post is that I would be working with digits on a screen, rather than having to read a scale, something which I have a problem with, sometimes.
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
Lord Kitchener
It depends on whether it's still parallel- the accuracy is not just for ease of fitting the parts together, it's also needed to sure that the opposite edges are parallel, so that the compononent has all four corners at right angles.
That shouldn't be a problem if you have an edge which is set precisely perpendicular to the saw to align the part against.
Quote:
Originally Posted by
Lord Kitchener
A decent random orbit sander is faster, and the discs can be changed quickly when they wear out.
Fair enough ... I thought you meant sanding by hand which would be slow!
I currently either clamp sheet material at the edges, or screw it to the bed with normal wood screws. Both methods hold it adequately and are much cheaper than a vacuum bed. The only real disadvantage is you have to leave tabs (or position the screws carefully) to stop parts you are cutting out moving. Don't be put off by the software - CAMbam free edition would do what you need and is (in my opinion) very user friendly.
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
Jonathan
That shouldn't be a problem if you have an edge which is set precisely perpendicular to the saw to align the part against.
Well, the movable fence is the reference edge, so if it doesn't move accurately, if one end moves a little further than the other, then it won't be parallel to the cutting line. An edge that was at right angles wouldn't help if it was a narrow piece, for instance 565 x 80
Quote:
Originally Posted by
Jonathan
I currently either clamp sheet material at the edges, or screw it to the bed with normal wood screws. Both methods hold it adequately and are much cheaper than a vacuum bed. The only real disadvantage is you have to leave tabs (or position the screws carefully) to stop parts you are cutting out moving. Don't be put off by the software - CAMbam free edition would do what you need and is (in my opinion) very user friendly.
Well, I am still considering it, so will bear that in mind, thanks.
Re: From Aylesbury, and looking for some help
Quote:
Originally Posted by
Lord Kitchener
Well, the movable fence is the reference edge, so if it doesn't move accurately, if one end moves a little further than the other, then it won't be parallel to the cutting line. An edge that was at right angles wouldn't help if it was a narrow piece, for instance 565 x 80
With a screw on both sides both ends will move the same amount .. near enough. True narrow pieces would be more tricky to trim accurately.
