So, I'm hopefully about to embark on the adventure of building my own CNC router :-)
I've tried to pick up as many tips as I can from the various build logs. Thanks to everyone who's posted info in them - reading them was very very useful! I hope I've arrived at some sensible design decisions but if anyone can spot anything that could be improved upon please do let me know - I'd very much appreciate your thoughts.
The machine will be predominantly used for hardwood. If it cuts ally then great but my priority is getting a good finish on wood first and foremost. The working volume I'm aiming for is 1220mm x 610mm x 220mm.
60mm x 60mm x 4mm steel box section, welded. Am thinking I'll use self levelling Epoxy on top before mounting the rails. I was considering making some 'bridges' across the structure which will be removed afterwards in order to level both sides up in the same pour. There will be a movable bed, as I doubt I'll make use of the full z travel very often - most of the time it'll be less than 100mm.
I've tried to take on board the comments people have made about getting the distances between the axis, linear bearings, ballscrews etc. as close as possible. For the Y, I'm thinking of a C shape made from extrusion, with the screw the 'right' side (I'll use bellows to keep the chips out).
Linear gear :
SBR20 supported round rail on all axis. RM1610 ballscrews driven by belts all round. X will be two screws driven by a single stepper. I'm thinking the Nema 23 3.1NM motors that everyone seems to recommend.
In my attached pics, green is 60mmx60mmx4mm box section steel, red is 90mmx45mm alu extrusion, orange is 60mmx30mm alu extrusion, blue is 20mm alu plate.
Would be great to get some feedback - thanks very much in advance!
me personally id put another leg in the middle either side of the table,other than that it looks great.
Cheers! Yeah makes sense - I did wonder if I should do that. Leg added!
Originally Posted by deisel
I would also consider some trusses (diagonal bracing) on the base in all axes to contain lateral movements rather than rely purely on the bending resistance of the box section
Thanks Irving, I'll look into that!
Originally Posted by irving2008
I understand it's customary that when one receives a parcel from Chai, a photo must be taken ;-) Got a week off work now to tackle the frame. Wish me luck!
Bit late to this thread has not been around much until lately.
Has you probably know I have lot of experience of steel framed machines of this size with mine being very similar. I see couple of issues you'll have with this frame and both will cause problems if not corrected.!! They actually relate to each other.?
The legs on each end need diagonal braces in both directions to support them, Esp the open ends.
If not done then they will flex and throw the ballscrew alignment off. You can't or shouldn't rely on the table to support them has every time you tighten or loosen the bed to move it the screw alignment will change and you'll get stalls and sticking. It only takes slight misalignment to severely affect performance or cause stalling at higher speeds when torques low.
This brings me to the next issue.! . . Ball-screw adjustment and alignment.? Make sure you build in some way to adjust for alignment. If the plan was to directly screw those FF _ FK bearings into the steel section then it will make it very difficult to adjust and align.
The screw alignment and setup is critical for best performance and can be frustrating but time well spent so make it easy on your self and think about adjustment, Both at the ends and on the gantry.!
Couple of other things I'll mention but not so important. The Rear plate on the Z axis doesn't really add much strength and unless perfectly machined square edges can cause binding on the bearings. It's also extra weight and expense you could drop.
You'll gain more strength by making a Z axis cover from 2-3mm Ali that fastens into the Z back plate sides and top bearing plate sides so bracing the 90deg angle between the two. You'll also have the added benefit it'll keep the shit out the Z axis bearings and screw.
This brings me next to Y axis belt sticking out the front.! It couldn't be in a worse place really has chips fly off the cutter in a linear direction bit like an hose pipe and it's in direct firing line so will get shot to bits.!!. . . It needs covering or moving.. . . The flying chips will get into it and wear it quickly.
Same with Rear belt but not quite has bad has it's higher and away from direct firing line.
Being honest with you your Gantry could be made stronger, neaten the Y axis belt arrangement and protect the ball-screw more if you adopt the L shaped gantry I've posted many times. Every thing is moved behind the firing line and also makes cable management easy it's all very neat. (tomorrow I'll post pics to show how neat)
Your using 90x45 anyway so it's an easy switch, the screw being moved back slightly more makes no difference to performance despite what certain others will try to tell you.
Sorry if I'm throwing a spanner in the works but some of these are important changes, the bracing is a must.!!
No worries about throwing a spanner - I really appreciate the advice!
1. Diagonal bracing makes perfect sense. Do you think triangular plates or more box section at a 45 degree angle would be better?
2. Ball screw alignment on X. I was planning to make oversized holes in the box section to allow for radial adjustment (ie. not threading into the steel but using a nut on the inside of the box section). I was thinking I'd put the gantry at one end, fix the bearing in place then move the gantry to the other end to fix that one. Do you think that would work? Do you also need adjustment in the axial direction somehow? I've actually got double fixed bearings for both X and Y (FK for X, BK for y) as I read somewhere that was better.
3. Ball screw alignment on Y. Here I was imagining I'd fix the double BK units to the 60x30 extrusion in the middle of the two 90x45s, then use oversize holes on the rails to do the same kind of thing as above, only this time allowing the rails to move slightly to make sure there's no binding. Does that make sense do you think?
4. Z axis rear plate. I'll scrap that in favour of some plates at the sides as you suggest.
5. Y axis belt protection / gantry style. Yeah, haha, funnily enough I'd had it the 'other' way around in all my designs up until the last revision. I changed after reading the thread where someone made a diagram showing the 'screw behind' approach wasn't as good. I get what you're saying though - it was definitely cleaner the other way around and kept everything out of the way. I'm kinda thinking I might switch it back.
6. Rear belt. I think I'll just put a board at the back to keep the chips away from this - seems easier than re-routing the belt.
Thanks again for your thoughts :-)
Either will work. just make sure you can triangulate the largest area you can. You really can't over do it but very easily can under do it.!!
Originally Posted by SimonC
Yes this will work fine. The Axial alignment needs to close but not super critical, the FK at each end will make it a bit trickier but you'll be fine. Regards if it's better or not then Yes but how much at this level is debatable and one I'm not getting into so don't anybody ask me too.!!
Originally Posted by SimonC
You won't fasten those blocks to the Extrusion slots of 60x30 if that's what you plan on doing.?
Originally Posted by SimonC
I may be miss understanding here.??. . . . Regards the rails and over size holes for aligning ball screws then that's a bit of hard way to go about it and will be pain. . . . . Over size holes and shims on the ball-screw mounts would be easier.
The most common cause of bent ball-screws is not from ball-screw end alignment but the distance from Ball-nut mount to screw centre. If it's out then when it approaches the bearing blocks it bends the screws. It's a slow job setting screw alignment and ball-nut mounting but worth the effort so building in has much adjustment potential greatly helps.
Well strictly speaking yes it's not the ideal but that depends on how far back and how it's done. The amount where talking here doesn't make a jot of difference and believe me if you built the machine your way and in the night I switched it without you noticing you wouldn't tell any difference, either in how it performed or quality of finish. Get silly about and then yes it show in cut quality and that's depending on what your cutting.
Originally Posted by SimonC
Last edited by JAZZCNC; 11-08-2013 at 09:19 PM.