584,826 active members*
4,984 visitors online*
Register for free
Login
IndustryArena Forum > WoodWorking Machines > DIY CNC Router Table Machines > My new CNC design - your opinion is needed! [RENDERINGS INSIDE!]
Page 1 of 3 123
Results 1 to 20 of 50
  1. #1
    Join Date
    Apr 2010
    Posts
    42

    Arrow My new CNC design - your opinion is needed! [RENDERINGS INSIDE!]

    First I have to thank the great community here in the zone!
    You helped me a lot. I started with nearly zero knowledge about cnc-buliding one year ago, reading through tons of treads here and I learned very much and finally I got addicted to it.
    Two months ago I thought I have collected enough information to design my own CNC-build.
    But I have some minor limitations on designing.
    -First is of course the price - I am still a stundent and money is a raw-material - you know.
    -Besides I am not very well equipped for machining, so i will have to reduce the complexity of my design to nearly basic. I just have normal homeworkertools and nearly no tools for metal machining. I will maybe gain some access at my school. They have tons of Mills and that stuff.
    And I got already the rails and blocks for free (nearly) from some friends and picked up an great Mitsubishi VFD on the bay.
    For the X-Axis I have two SIZE 20 1390mm Bosch-Rexroth rails
    The Y are the same but SIZE 15 and 1140mm long and on the Z I got some pretty completely new and sealed THK rails (450mm long).



    The idea to my concept:

    Regarding to my situation the size of my build was already defined by the rails I got.
    for this reason it has to be a relatively big machine design. So my first idea with an MDF/Wood structure died because of my requirement to mill some Aluminium or maybe small amounts of Steel and the high resulting forces applied to the long gantry.
    So I chose Aluminium. - The sides, some of the base framing and the Z-Axis will be made out of watercutted 20mm Aluminium sheets. As far as I know that is the easiest sollution for my lack of tools. (note: lasercutting is nearly impossible for that thickness and miling would be too expensive for me)
    The other structures will be made with aluminium extrusions (30x60 for the base and 30x120 for the Y-Axis)
    I chose a dual X-Axis concept because it reduces the deflection from the gantry and I will be also able to make the base more massive. I gonna drive it with two stepper motors an each side because I am afraid that the so extremely long timing belt that would be required when only one stepper in the middle would stretch too much resulting to a high backlash.
    But all steppers are belt reduced. That lets me choose if I want double the resolution (2:1) or the double the speed (1:2) or let it genuine by 1:1 simply by changeing the pulleys or/and buying 4 additional belts for 1:1. Belt width will be 9mm or more.(would you prefer HDT 3mm?)

    SCREWS:

    All screws will be ballscrews (class7) with 5mm pitch (20mm dia and 16mm for Z). At the moment I think I will buy them from chai because of his very low price (I have already read the stories about him). Here in Austria I dont think that I will get a ballscrew lower than at least ten times that price. (also C7)

    DRIVE:

    For the drives I chose 3NM stepper (NEMA 23- bipolar parallel setup to 2.8AmppP)
    I will get them for a fair price from BZT in Germany.

    DRIVER:

    I am going to use a Gecko540 because of its really great quality compared to the price. I think I cant do much false when i take it.

    SPINDLE:

    One very important part to me is choosing the right spindle motor. I think it is an very common problem to many DIY-CNC-builders. Most try to save money by taking a Kress spindle or something like that and let that poor device mill some aluminium until its bearings are dead. Others invest in these cheap china spindels which are pure crap for metal milling. I have some connections to some manufacturers and nearly all trustable are saying that they are only designed for wood engraving (noway metal!).
    But I found a chinese spindle which is designed for milling and engraving soft metals. It is equipped with 4 quality nsk bearings and has a power input of 1.2kW. Compared the the cheap china spindles it has 0.96 Nm torque (cheap 1.5kW ones have only around 0.6Nm. and it weights about twice as much (8kg) It has an ER20 collet and it is water cooled. But that sucker is only being shipped for around 400€ - anyway it seems to be a great deal - and i dont want to save money on the wrong place.

    My goal is to gain an accuarcy of about 0.05mm per 300mm (including deflection, pitch error and run out of course) Pitch errors can be a little bit compensated by Mach 3 - so C7 ball screws arent not a too big problem
    As said before I will mill aluminium and some steel but also PCBs, wood and huge foam blocks (ex .:for molds).
    Maybe I could later make some money by firing it up for Milljobs in the architecture and hobby-aircraft sector.

    My first costlimit was 1000€ but with my needs and my actual design I am FAR away from that.
    I was really shocked when I saw the price for an 1000x1000 20mm aluminium sheet (560€). I have calculated the sheet surface i need for my parts. it is 0.86m² (so around 450€ on sheet material only). I have no plan how expensive the watercut will be - anyway I am afraid from that.

    ...enough text...HERE are the renderings of my first design:
    (more information is blow them)






    You may have detected that my Z-Axis is lookig a bit different than regular designs. I really prefer it, because it makes the Z-Plate much more stiffer and gives me a bit more travel. The only con is that the Z-Stepper has a bit more work to do, because he always have to lift itself and the rails. But I think I can live with that. That method seems to be a tick easier and also saves me a little bit of aluminium sheet.

    At the moment the spindle can travel 900mm on Y, 1150mm on the X and 220mm on Z. That is more than I would ever need, but I dont want to cut my already existing rails down. I have already increased the distances between each pair of blocks - to reduce the momentum (torque I guess in English)

    I will use just a simple MDF board as a working surface for my first couple of tests. It doesnt hurt my wallet if the spindle goes to much down the Z. Maybe I will change to an Aluminium T-Slot surface later then. For Metal milling I think that I will remove the MDF board and mount the work pieces directly on the base extrusion framing (you can see it in the 2nd picture)

    So that is all I have to say first - now you have to tell me what I should improve. If you need any plans, please send me an PM with you e-mail adress. I can send them as an e-drawing 3D file.

    By the way: what do you think about "BLOCKSTROYER" as the name for that machine? (honestly)

    A website with the coming build progress will be released soon, so please look back later for further information.

    I hope my English isn´t too confusing for you! Please feel free to correct me. (PM)

    grz. Max!

  2. #2
    Join Date
    Aug 2008
    Posts
    409
    looks good, a couple points..

    Your Z ballscrew will need a brake to keep it from dropping when the drives are disabled.

    Having your Z motor mounted to the moving plate complicates the wiring a bit.
    I would recommend you do something to cover up the Z motor and belt and something to cover up those X axis belts, you dont want any debris or anything getting in between the belts and pulleys.

    Dont expect this machine to be able to machine steel, it may even have a hard time getting a good finish on Aluminum. I wouldnt even consider doing mill jobs for aircraft parts, that would require some pretty high tolerances.

    Also expect to pay 2-3x that 1000 budget.

    What size working area are you going for? I may have missed it.
    Otherwise it looks pretty good, nice work on the renderings. Looks like you spent a lot of time designing.

    Should be a nice machine!

  3. #3
    Join Date
    Jan 2007
    Posts
    723
    You will probably get better performance if you dont use the belts and drive the screws directly with flexible couplings. One of my machines uses belts with servo drives and I can see the inaccuracies and flex that the belts cause. My first machine users direct drives with flexible couplings and it is rock solid with better torque and accuracy.
    http://www.glenspeymillworks.com Techno LC4896 - 2.2Kw Water Cooled Spindle | Moving Table Mill from Omis 3 CMM, 500Lb granite base | Epilog Legend 32 Laser Engraver

  4. #4
    Join Date
    Apr 2010
    Posts
    42
    thanks for your replys! You really let me think about by choose on the timing belts. I already thought about direct drive to ball screw - but I was very unhappy, that I will never be able to change the resolution of my machine.
    But with a maximium Stepper torque of 3Nm I will have with a 30mm pulley dia a force of 200N pulling on the timing belt (3Nm:0.015m=200N)
    But I dont know the modulus of elasticity of the timing belts i would choose. (They are called SynchronFlex and are advertised to have a bit more strenth due more glass fibres) - So I cant calculate the actual flex of the timing belt and regarding to that the resulting inaccuracy

    Anyway I already thought about covering up at least the X-Axis belts and pulleys. - Maybe I will cut some custom housings out of MDF when the Machine is up and running.

    But apart from that I am very unsure that I should bet on ballscrews.
    Many of the builders here in the zone are really promoting a R&P system.
    But I dont know where I can buy them from and how expensive they would be.
    Mainly I think that they will wear of much more faster than a ballscrew. When just a small piece of dirt gets on the rack, will result an false aligenment.
    Will they be the same accurate as a ball screw would be for the same price? That does only matter for me. Why should I make it more expensive when it is useless(?). So please help me.

    On little thing I NEVER understood:
    Why the heck are all driect driven machines coupled with a flixible coupler to the stepper?! This is just an backlash that goes slowly to zero when lowering the actual torque on the coupling right? But how could that help me? For example: When I let it cut an simple circle it will be deformed after each point of return(?).

    Does anyone of you can tell me how much forcees are working axially by working for ex. Aluminium or steel. (for sure I would need a formula for calculation) Where can I learn how to do that? (any links?)

    So then I will be able to calculate how good my design will be for milling aluminium and how much it is going to flex and the resulting inaccuracy will be.
    @Phife: You are pretty colse with your appraisal ;-)

    I need still more opinions!

  5. #5
    Join Date
    Feb 2009
    Posts
    2143
    Quote Originally Posted by MaxTheModder View Post
    On little thing I NEVER understood:
    Why the heck are all driect driven machines coupled with a flixible coupler to the stepper?! This is just an backlash that goes slowly to zero when lowering the actual torque on the coupling right? But how could that help me? For example: When I let it cut an simple circle it will be deformed after each point of return(?).
    As much as we all would like it, nothing is perfect. If you bolt the stepper directly to the screw, there is NO room for error. Since no shaft spins on a perfect center, this will rapidly lead to wear on the bearings in the stepper. You must have some room for compliance. In the design you show, that compliance is the belts. You can make zero-backlash couplers cheaply that are zero-backlash. But SOMEWHERE you need to take up "real world" inaccuracies or your machine won't run for long!

  6. #6
    Join Date
    Apr 2010
    Posts
    42
    Ok - so the flexible coupling does just smooth down the screws rotatioan to keep the steppers bearings longer up? But the screw is already mounted in an angular contact bearing before the stepper. So I think there would be even less radial forces as a timing belt and a pulley (it needs some radial force to be tight) would ever make. - so far as I can imagine.

  7. #7
    Join Date
    Feb 2009
    Posts
    2143
    The flexible coupling allows the motor and the screw to be mounted and not be PERFECTLY aligned to one another, which is impossible. Some couplings allow up to 15 degrees of misalignment.

  8. #8
    Join Date
    Apr 2009
    Posts
    5516
    I use belt and pulley to drive two screws in my machine. Although I could use a larger motor, it runs fine and I get good repeatability.

    Check out some large knee mill conversions to CNC. They have belt drive, and look at the weight they have to move!

  9. #9
    Join Date
    Apr 2010
    Posts
    42
    so here is a little update:
    I designed the cover for the timing belts on the X axis.
    Its going to be a 30mm MDF custom milled to fit with that machine.



    I still have to design a cover for the X-ballscrews. But I think that would also be no problem.

    I have also calulated a little bit:

    regarding to avoid any resonance on the screws the fastest possible speed will be
    82mm/s on X (3300pps Stepper 2:1 reduction)
    133mm/s on Y (5333pps Stepper 2:1 reduction)
    and resonance maximum on Z is 800mm/s but will be jogging around like Y

    Maximum forces that can be applied: 12000N on each axis (twice on X) (with 3NM stepper force and 2:1 reduction - is just holdingforce - calculated with 90% efficiency factor on screw, and disregarding loss on belts)

    But with that force the Y´s porfiles are going to defelect too much ~ I think it will be at around 1-3mm defelection in X

    But I think that I will never need 12kN so maybe I will choose a smaller stepper.
    (when in move force is going to cut down to about 3000-6000N per Axis)

    thats it for today - any opinions are welcome!

  10. #10
    Join Date
    Apr 2009
    Posts
    5516
    Quote Originally Posted by MaxTheModder View Post
    so here is a little update:
    I designed the cover for the timing belts on the X axis.
    Its going to be a 30mm MDF custom milled to fit with that machine.



    I still have to design a cover for the X-ballscrews. But I think that would also be no problem.

    I have also calulated a little bit:

    regarding to avoid any resonance on the screws the fastest possible speed will be
    82mm/s on X (3300pps Stepper 2:1 reduction)
    133mm/s on Y (5333pps Stepper 2:1 reduction)
    and resonance maximum on Z is 800mm/s but will be jogging around like Y

    Maximum forces that can be applied: 12000N on each axis (twice on X) (with 3NM stepper force and 2:1 reduction - is just holdingforce - calculated with 90% efficiency factor on screw, and disregarding loss on belts)

    But with that force the Y´s porfiles are going to defelect too much ~ I think it will be at around 1-3mm defelection in X

    But I think that I will never need 12kN so maybe I will choose a smaller stepper.
    (when in move force is going to cut down to about 3000-6000N per Axis)

    thats it for today - any opinions are welcome!
    The other way around it, is to use a screw with high lead, thus decreasing the rpm's on the screw. You'd have to use larger steppers or servos to overcome backdriving in x and y in addition to gearing down, and you could use a standard pitch in z axis.

  11. #11
    Join Date
    Aug 2008
    Posts
    77
    Hi

    Did not read through the whole thread, so maybe it's been mentioned...
    From the simulations I did, I found the the lower Y bearing takes the most force from the Z axis. Maybe put another piece to form a T?

    Cheers,

  12. #12
    Join Date
    Apr 2010
    Posts
    42
    yes- I did already noticed that the lower Y 120x30 profile takes the most force in X direction. But the sloped gantry shape doesnt allow that I change the orientation of the lower and upper extrusion by 90°.
    I finally came to the solution, that I will try it in this configuration and if it will deflect too much, I will get some U-shaped steel profile and put it on the back of the lower Y-extrusion

    @louieatienza:

    I am sorry, but I dont know what you mean.
    I my mind with 5mm pitch - the nut will got 5mm with one revolution.
    When I get a a higher lead (=pitch - right?)I will loose half the resolution
    and need as you said twice the force on the steppers to apply the same holdingforce.
    At the moment I am happy with 80mm/s speed on X. I think I will never go faster when milling so it doesnt matter for me.
    but with 12kN holdingforce I dont think that it would backdrive anyway (disregarding lots of deflection ;-)

  13. #13
    Join Date
    May 2005
    Posts
    1662
    An axis with a 5mm lead ballscrew has reasonable resistance to movement even with no power to the steppers. When powered up it won't be a problem. I have no experience with higher lead screws.

    Timing belts can be found on commercial machining centers. Size them appropriately and they should work well.

    Nice renders.

    My 2 cents fwiw.
    Anyone who says "It only goes together one way" has no imagination.

  14. #14
    Join Date
    Apr 2009
    Posts
    5516
    Quote Originally Posted by MaxTheModder View Post
    yes- I did already noticed that the lower Y 120x30 profile takes the most force in X direction. But the sloped gantry shape doesnt allow that I change the orientation of the lower and upper extrusion by 90°.
    I finally came to the solution, that I will try it in this configuration and if it will deflect too much, I will get some U-shaped steel profile and put it on the back of the lower Y-extrusion

    @louieatienza:

    I am sorry, but I dont know what you mean.
    I my mind with 5mm pitch - the nut will got 5mm with one revolution.
    When I get a a higher lead (=pitch - right?)I will loose half the resolution
    and need as you said twice the force on the steppers to apply the same holdingforce.
    At the moment I am happy with 80mm/s speed on X. I think I will never go faster when milling so it doesnt matter for me.
    but with 12kN holdingforce I dont think that it would backdrive anyway (disregarding lots of deflection ;-)
    That's not the point. 80mm/s = 4800mm/min. For you to achive that speed your screw will have to spin at about 9600rpm, which is probably over the theoretical limit for your leadscrew at that length! I don't know the torque curve of your steppers, but I'd check to see that you are getting 2.4Nm of torque at 9600rpm! And that's unloaded speed without the friction of your bearings and weight of your gantry! To gain back the resolution, however would need gear reduction, but that would also bring the torque back up....

  15. #15
    Join Date
    Apr 2010
    Posts
    42
    @louieatienza
    I dont know how you calculated that RPM - but with my calculation I come to 960rpm at 80mm/s jog 990rpm is the maximum for resonance (based on the nookindustires calculator). With 2:1 belt reduction that is pretty close 2000 rpm on stepper (6600PPS) that is going to cut down the torque by about 50-80%. For this reason there should be at least 2-3Nm torque on the screw that results an maximum force of 3000-6000N per axis


    At the moment, I am working on a new gantry shape that is about 1.5 times mores sturdy as the current shape with even less mass. I simply change the orientation of the extrusion by 90° - it looks a bit funny but I think I gonna get way better results

    check back in approx 2-3 hours for the new renderings!

  16. #16
    Join Date
    Apr 2009
    Posts
    5516
    Quote Originally Posted by MaxTheModder View Post
    @louieatienza
    I dont know how you calculated that RPM - but with my calculation I come to 960rpm at 80mm/s jog 990rpm is the maximum for resonance (based on the nookindustires calculator). With 2:1 belt reduction that is pretty close 2000 rpm on stepper (6600PPS) that is going to cut down the torque by about 50-80%. For this reason there should be at least 2-3Nm torque on the screw that results an maximum force of 3000-6000N per axis


    At the moment, I am working on a new gantry shape that is about 1.5 times mores sturdy as the current shape with even less mass. I simply change the orientation of the extrusion by 90° - it looks a bit funny but I think I gonna get way better results

    check back in approx 2-3 hours for the new renderings!
    Yes I meant to type 960, a typo on my part. I don't know why my keyboard doubles the 0's (it double the v's as well)... Which brand stepper are you using?

  17. #17
    Join Date
    Apr 2010
    Posts
    42
    I think I will use BZT steppers (german brand). I know they are not produced by themselves, but I think they will not sell crap when thay have to take real 2 year warranty.
    At bipolar prallel they have a almost linear torque curve.

  18. #18
    Join Date
    Apr 2010
    Posts
    42
    so here is the new gantry design:





    As you can see in the last pic - the lower profile is now turned by 90° to better handle the force. Its about 2 times higher as at the upper profile in X direction
    so I think it is way better and more sturdy as the privious design.

    what do you think?
    I gonna work on the full new updated design if this is better

  19. #19
    Join Date
    Apr 2009
    Posts
    5516
    Quote Originally Posted by MaxTheModder View Post
    so here is the new gantry design:





    As you can see in the last pic - the lower profile is now turned by 90° to better handle the force. Its about 2 times higher as at the upper profile in X direction
    so I think it is way better and more sturdy as the privious design.

    what do you think?
    I gonna work on the full new updated design if this is better
    A great place for the cable carrier as well!

    Back to the steppers, I don't lnow many brands that perform well at 1000rpm, let alone 2000rpm. I'm interested in purchasing some as well, if you could direct me to your source..

  20. #20
    Join Date
    Apr 2010
    Posts
    42
    here is the stepper:
    eBay Österreich: Stepmotor Schrittmotor 4,2A 3Nm CNC Fräse Fräsmaschine (Artikel 270712289529 endet 08.03.11 13:41:39 MEZ)
    It is in german at all, but the graph speaks for its own.

    Yeah the lower extrosion is a really welcome place for cable chain.
    But that is not very independent at the moment. - I just want to optimize its performance.

    Can someone tell me how I can calculate the cutting force?

Page 1 of 3 123

Similar Threads

  1. Another design question... need opinion.
    By SatanKlawz in forum DIY CNC Router Table Machines
    Replies: 0
    Last Post: 02-07-2010, 11:54 PM
  2. Opinion on Base Design
    By Mr.Chips in forum DIY CNC Router Table Machines
    Replies: 11
    Last Post: 01-21-2009, 01:33 AM
  3. modern commercial cnc design books / inside details
    By bobJandal in forum Uncategorised MetalWorking Machines
    Replies: 6
    Last Post: 11-01-2008, 12:21 AM
  4. opinion needed
    By mm12348 in forum Uncategorised MetalWorking Machines
    Replies: 0
    Last Post: 04-26-2007, 02:58 AM
  5. what coolant to use when milling copper?(3d design inside)
    By HawainPand in forum MetalWork Discussion
    Replies: 14
    Last Post: 04-18-2007, 03:51 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •