Milli a new composite mill kit - Page 2

Page 2 of 105 FirstFirst 123451252102 ... LastLast
Results 21 to 40 of 2096

Thread: Milli a new composite mill kit

  1. #21
    Member
    Join Date
    May 2016
    Location
    United States
    Posts
    316
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Following with interest. I really hate saddles, too.

    Mill No5 - X-servos could be mounted behind the column. Cast holes for the ballscrews and mount the motor and bearing block in back. 'Outboard' screw block up front. Saves a bit of space up front and protects the motors..

    You might even get away with no outboard ballscrew bearings at the front if you up-size the screws enough and don't expect really fast rapids on X. Mounting the ball nut at the back of the table means you could get full travel and the ballscrew ends would never be exposed.

    Downside is that you'd have no mold draft where the bearing blocks mount... maybe not a problem if you're just going to break the mold apart after the cast.



  2. #22
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi Spumco - Yes I'm looking at rear mounted motors (or reflex them with belts) and other ways to combine spaces now the general layout is done. This is intended as a production machine so won't be "breaking" moulds. Moulds will be fibreglass. I'd like to use Hiwin linear drives but they haven't got back to me with prices I made a patient lifter in a hospital once and left the screw end free. After being used for a while they decided they wanted it to lift faster so I cranked up the inverter a bit and the free end wobbled terribly and the nut whined in unison Luckily there was a spot for a support bearing. So I'll have bearings both ends as good practice... Peter



  3. #23
    Member Sterob's Avatar
    Join Date
    Jan 2007
    Location
    Australia
    Posts
    626
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Mill No 5:
    3 rails and 2 ballscrews to align on the table. Fairly complex.....Will they be more difficult to align?

    Steve



  4. #24
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi Sterob - 3 rails yes I suppose more work to set up then two but the table will be long so its either a very thick table or more bearings. I choose more bearings... I'm sure if you can set up the two outers like normal then the middle one is easy. Two drive screws just the same as a moving gantry router (In fact easier as the table can't walk) so I think its a good compromise maybe a plus, until its working won't know. Can't see any show stoppers or blue screens yet... Peter



  5. #25
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi All- The thing to sort now is the spindle or power head and motor. BT30 power head - seems to be lots of these around. I'll look up some mill speed and torque specs and set these up before I start detailing... Peter



  6. #26
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi All - This may seem an odd thought but why aren't steppers used for spindle drives? is it the intermittent motion? or the speed range? A 1000W servo can do 3Nm continuous so can a stepper? The servo will have more speed range and the torque curve is flatter. Say we need 1000-5000rpm and 1000W ? The stepper torque curve is poor for the application I think...

    Taig has 187W 1000-11000rpm
    seig have 500W 100-7000rpm
    grizzly has 750W 50-2250rpm
    haas mini 6.5HP 6000rpm vector drive

    what speed is needed for rigid taping or what is the ideal specs for a small mill motor? This is an area I'm not familiar with... Peter

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-1000w-servo-jpg  


  7. #27
    Member
    Join Date
    Jan 2008
    Location
    Australia
    Posts
    1523
    Downloads
    2
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Stepper torque drops off badly with rpm

    Modern machining methods use higher rpms. Carbide uses higher rpms/surface speed



  8. #28
    Member
    Join Date
    May 2016
    Location
    United States
    Posts
    316
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Steppers are used extensively as "spindle drives" in applications where the load requirement is constant. Peristatic metering pumps are a good example. Load is known, so the stepper can be sized appropriately. Benefits are very precise metering and no PID tuning or servo dithering.

    Where steppers fall short of PM servos or induction motors is the narrow torque curve (as Pippen noted above), as well as very little reserve torque in overload situations. I believe some of the newer closed-loop stepper drives can increase the available torque a little if they detect a following error, but nowhere near the huge short-duration overload capacity of modern PM servos.

    And you really can't get anything bigger than a NEMA52 stepper - and those have extremely narrow (but huge) torque curves.

    Rigid tapping depends on the material, tap size, thread pitch, etc.

    Spindle torque/speed requirement also depends on intended machining tools, materials, ad nauseum. It also depends greatly on the rigidity of the machine... a small hobby mill simply can't take advantage of high-horsepower with milling; it's just too flexible to remove material aggressively. Perhaps while drilling or tapping, but any side loads with a big cutter and lots of power will result in chatter and broken tools.

    Tormach has a whitepaper on smaller machine spindle power requirements - shoudl be available on their site if you dig around a bit.

    The best of all worlds is to have a transmission system so you can have low speed and very high torque for big drills, taps, and face mills... but also high speed for small cutters. If the spindle encoder is directly on the spindle (and not just on the motor), then you are not limited to timing belts for spindle orientation during tapping or other operations sensitive to rotational position (or backlash).

    Options:
    1. Spend a fortune on a really high-end spindle motor that can 'do it all'
    2. Compromise on a spindle motor with a single belt reduction ratio that permits acceptable torque down low and meets the upper RPM target the spindle can handle.
    3. Use a geared-head spindle. See all the medium-sized benchtop mills and similar.
    4. Use a two (or more) speed belt reduction transmission. Perhaps 2:1 and 1:2.
    4a. Manual belt changes; very common. See Tormach, Bridgeports, and others.
    4b. Automatic belt change. See Fadal's version. Very slick and a little complicated.

    Link to video of a guy from Germany who built a Fadal-style belt changer. Rather clever of him.




  9. #29
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi All - I have "talked" briefly to DMM and and their DST 130mm frame seems to be the go. The 130mm frame has 3 power options 1.0, 1.3 and 1.8kW. So the owner could choose or upgrade as needed. I looked at the Nema42 which I like as its a smaller frame. But then upgrading means more bracketry. So onward to detailing the motor mount.... As the motor will have a belt drive the owner can pick the drive ratio to best suit their work. Peter





  10. #30
    Member
    Join Date
    Jan 2008
    Location
    Australia
    Posts
    1523
    Downloads
    2
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Peter,

    My proposed mill design is essentially very similar to the Vulcan mill you posted in another thread (https://www.cnczone.com/forums/verti...ml#post2400702).

    Single ball screw on each axis - easier tuning / ball screw error compensation

    I have 45mm roller linear bearings. Completely oversized but stiff and I got them cheap. I'll have rather large bearing spacing.

    My mill is designed as a one off around some overkill roller linear bearings and ball screws I got cheap. In my case, as a one off, going for over spec rather than fine engineering is more time efficient.




    Let's look at the rough pros and cons of each basic mill design:
    Eumach - Vertical Horizontal Machining Center / Taiwan Manufacturer has a lot of the following designs with photos of the frames.

    C Frame
    Pros
    - probably easiest to make
    - Z only has high as needed
    Cons
    - large head overhang (Y travel) reduces stiffness
    - saddle can be difficult

    Example:
    Vertical Machining Center / VMC-1100 / Eumach

    Moving Column / Moving C frame with Moving Table
    Table moves in Y, column moves in X, head moves up and down column (Z)
    Pros
    - no saddle required
    Cons
    - large head overhang (Y travel) reduces stiffness
    - moving the column mass


    Moving Column / Moving C frame with Fixed Table
    Table fixed. Column moves in X and Y, head moves up and down column in Z
    Pros
    - good for very heavy work pieces
    Cons
    - large head overhang (Y travel) reduces stiffness
    - moving the column mass

    Example: Traveling Column Machining Center / TVMC-2000 / Eumach

    Rising Gantry / Portal Mill
    Pros
    - Stiff design - Z only as high as needed
    - Can have large Y travel without compromising stiffness
    - Can have large X travel without compromising stiffness much
    Cons
    - dual drives on columns OR central screw for gantry rising is somewhat complicated

    Example: Vulcan Machine Co - still under development
    https://www.vulcanmachineco.com/
    Despite considering this design myself, I do have some concern that the big VMC companies are not using this style...

    Fixed gantry
    Pros
    - stiff than C frame due to dual columns / 'closed' frame
    - single screw / drive on every axis
    Cons
    - large Z overhang (Z travel) reduces stiffness

    Example: Okuma M460 - regarded as very stiff machine
    https://www.okuma.co.jp/english/product/vmc/genosm.html


    Moving Gantry
    Pros:
    - Largest work area for overall machine size
    Cons:
    - Reduced stiffness
    - large moving mass (can reduce this by having high fixed sides so no columns to move)
    - requires dual drives

    Example: https://hu.dmgmori.com/products/mach...dmu-340-gantry


    Horizontal Machining Center
    Not in my use case and not considered here

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-bridge-okuma-4-jpg   Milli a new composite mill kit-c-framemill-design-jpg   Milli a new composite mill kit-okuma-m560v-x-bracing-png  


  11. #31
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi Pippin - There's probably not a magic bullet design that solves all the issues to milling something. In 10 years time mills may not exist as we know it. They will be 3D printers with finish machining capability. But there is one idea I will look at before getting serious with No5 that I have not seen before. Will take a few days to get it modeled to the same detail as No5 so I can compare apples to apples. Then I'll get on with detailing. Its an extension of No5 to remove the cantilever aspect of the column. I've been looking at DMM servos and I think the 1kW is overkill for this mill. I think 750W 1HP is fine for a starter mill. Unless the 1kW and 750W are similar prices but I think not. Waiting for some budget prices to come through. I'm excited about this build. The mouldings make the part count so much less and I have control over geometry that is impossible with the sheetmetal approach. Only hassle is building the moulds but once built can make as many machines as needed.... and does UCCNC run servos? I'll have a look in the manual... maybe time to move up to servos for motion...Keep at it Peter

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-dmm-motors-jpg  


  12. #32
    Member
    Join Date
    May 2016
    Location
    United States
    Posts
    316
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Quote Originally Posted by peteeng View Post
    and does UCCNC run servos?
    Short version: Yes it does.

    Long version:

    UCCNC is a step & direction axis output controller. Depending on the hardware controller (UC400ETH, UC300ETH, AXBB-E) it can output up to 400kHz pulses per axis. No issues controlling DMM servos (or any other drives which accept S&D inputs).

    UCCNC can use PWM, analog, or S&D for spindle control. The DYN4 drives can accept all three of these input modes.

    If set up as S&D, the spindle can also be programmed in UCCNC to be an axis using the same pins. This might be a slight bug, but can be exploited to control the spindle servo with common "S" word G-code commands (for standard RPM) and do spindle positioning for homing/ATC alignment or single-point broaching.

    Example would be to edit the "M6" macro inside UCCNC so that the spindle stops, then the C axis is commanded to home to a proximity switch (or the DYN4 Z-pulse if using a 1:1 or 1:2 drive ratio) so drive dogs line up.

    Breakout boards plays a significant role, as cheaper BOB's have 100 or 200kHz optocouplers. This limits the resolution or speed possible on an axis (or S&D spindle). CNCDrive's UCBB and UCSB have high-speed optos which will not limit the output frequency.

    As with all other PC-based hardware motion controllers, the MC buffers the signals from the PC to create very steady output signals. If one is used to real-time industrial controllers it takes a little getting used to the (very) slight lag between commanded inputs and the resulting output. Pushing the feed hold button on a true real-time industrial controller means INSTANT feed hold. Same thing on UCCNC (or Mach3 using a ethernet smooth stepper, or any other PC with hardware motion controller) results in a very slight delay before the feed hold while the buffer empties before the next command is processed. It's subtle, but perceptible. Buffered signals = smooth, but also delays.

    UCCNC spindle control considerations:
    - Encoder input frequency is limited. Fairly coarse encoder counts are required so as not to saturate the controller. DMM DYN4 encoder outputs can be adjusted to 2000 pulses per rev minimum using the LINE_NUM parameter. This means at 3kRPM it will be outputting 100kHz signals to UCCNC. This is well within the usual optocoupler speed, but may be too fast for the controller or UCCNC. Suggest contacting CNCDrive directly on this issue.
    - Encoder input signals must be single-ended (no differential). Possible noise issues over longer distances unless a differential-to-single converter is used.
    - S&D spindle control mode can be very abrupt start/stop. Acceleration must be set experimentally as well as adjusted in the servo drive to avoid tripping the drive. Braking resistor essentially mandatory.
    - Spindle speed PID is available and works well (closed loop spindle)
    - Spindle gear ratio changes possible on the fly (via M-codes) (i.e. you physically change from high to low speed and easily select new ratio in UCCNC so it calculates the actual spindle speed properly)

    UCCNC axis control considerations:
    - No analog or PWM axis output (i.e. torque mode servo)
    - No CAN, Ethercat, or Modubs plugins exist as far as I know (but might be possible)
    - No axis feedback loop to the motion planner. You can connect the servo drive encoder outputs (or linear scales) to UCCNC and can read/display the encoder counts, but the motion planner does not use the data to adjust trajectory.

    I've been using UCCNC for 4 years now on multiple machines and am happy with it. It has limitations, but my opinion is that it's the best of the non-high-end CNC controllers out there. Other controllers offer more features (closed loop trajectory, lathe features, etc), but they are either extremely expensive or have an extremely steep learning curve (or both). There is a robust community on the CNCDrive forum who are quite helpful, and CNCDrive themselves are very responsive to technical questions or assistance.

    As for DMM servos, they're fine. There are cheaper servos with the same basic specs available from China, but support is non-existent. There are other Chinese servos with better specs and great support (Delta), but they're more expensive. And there are non-Chinese servos out there with really great features that are insanely expensive.

    DMM seems to have hit a sweet spot between features/specs/support and price.

    Minor irritations with DMM servos, based on my limited experience:
    - Drives are extremely tall (deep) compared to most other drives I've played with. They don't fit inside a standard 8" deep enclosure - the DB25 I/O cables stick out too far and you have to get a stupidly deep box to fit them. (But they're also narrow, so that's nice for mounting.)
    - Motor power/encoder cable connectors are terrible. They are not waterproof and there is no strain relief. This can be fixed by chopping off the connectors and replacing with higher-quality ones but it's something to consider (maybe $30US per motor)
    - Encoder signals are a proprietary serial signal. Cannot use the motors with an 'agnostic' servo drive which looks for standard quad A/B/Z inputs.
    - Encoder cable at drive end uses a really dumb IEEE 1394 (Firewire) connector. Good luck adjusting or modifying the cable at that end to suit a particular enclosure.
    - DMM's tuning software is OK, but it's missing a simple 'jog' function. You cannot just hold down a jog key to move the motor a bit; you have to command it to move a specific number of encoder counts or a continuous RPM. Minor complaint, but really aggravating when you want to just jog it a bit during commissioning.
    - Cables are inexpensive, so that's a plus.

    Regarding servos in general...If you are building a single mill for yourself there are other options which permit superior features at lower cost than DMM (ebay drive/motor combinations). But not recommended if you intend to build something for someone else as the combination will be 'weird' and possbily hard/expensive to get replacements. If you want easy to use, and easy to replace, then DMM is hard to beat.

    Wrapping up a 12'x6' plasma build, using UCCNC, UC300ETH, UB1 BOB, and DMM DYN4 + 750W servos. 1200+IPM rapids with <0.001" X/Y resolution.

    Yes... UCCNC can control servos quite nicely





  13. #33
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi Spumco - Thanks for the "summary". All my machines are designed as production machines so need a consistent source for parts. So DMM is looking good. I'll have to read your summary a few times for it to sink in Ta Peter

    couldn't get your link to work....



  14. #34
    Member
    Join Date
    May 2016
    Location
    United States
    Posts
    316
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Which link?



  15. #35
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    this one - Peter

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-link-jpg  


  16. #36
    Member
    Join Date
    May 2016
    Location
    United States
    Posts
    316
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Strange. I just uploaded a couple photos of a UCCNC/DMM machine I'm about done building - should have been embedded in the post and not linked.

    I'll try again...

    Milli a new composite mill kit-20201009_223727-jpgMilli a new composite mill kit-20200105_232126-jpg

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-20201009_223727-jpg   Milli a new composite mill kit-20200105_232126-jpg  


  17. #37
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi Spumco - Very nice work. Very tidy wiring, I think your screwdrivers could be better organised to match... Can't live without forceps. Just finished the wiring diagrams for the sparky for my shed, looks like your box just 9mx14m...lights, GPOs switches, subboards, very much the same.... Now back to checking YaGs motion parts so I can get those on order today. Peter

    Oh by the way Stephen at DMM has been very responsive and helpful...

    DMM | AC SERVO DRIVE | AC SERVO MOTOR | ROTARY ENCODER

    see my style of wiring - I'll use a box for this mill.... This is Brevis No2 it trials geared steppers (N17s with 10:1 so strong!) and running a laser (not yet right)

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-shed-lights-jpg   Milli a new composite mill kit-shed-gpos-jpg   Milli a new composite mill kit-electronics-jpg  
    Last edited by peteeng; 10-24-2020 at 10:44 PM.


  18. #38
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hello all you cnc lurkers out there;

    I've started Milli No6. It's a combo of ideas. There are machines with rear "walls" and double columns like the vulcan but they usually have twin high rails, a saddle and a Z axis. The No6 wall makes it very stiff. Since this is a benchtop machine this wall also helps with keeping the muck contained. The 3 rail moving table gave me the idea to extend this to the Z axis. Plus when looking at the FE deflection the column and or the table is the bendy bit so looked at supporting the top of the column somehow. I suppose the "gantry" or the vertical bridge or column is now a saddle. But a supported beam is much stiffer then a cantilever and also its vibration modes are less.

    Things to do:

    I have found a supplier of the motor mounts that does not have the little step in it. I think this will make moulds easier and in this design I can put the table motors at the rear. ( Mactec)

    It has the same kinematics and envelope as No5 so I can do an apple vs apples FE soon... So flesh it out a bit more, All comments welcomed.. I think I'll call it a tombstone or headstone config...Peter

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-milli-no6-jpg  


  19. #39
    Member peteeng's Avatar
    Join Date
    Jul 2018
    Location
    dum dum
    Posts
    6248
    Downloads
    0
    Uploads
    0

    Default Re: Milli a new composite mill kit

    Hi All - Its magic - I organised No5 and No6 into the same height and depth. I then made all parts steel to represent the max stiffness it could be. No5 weighs 1.26 tonne and No6 weighs 1.77 tonnes. I bonded these all together and pulled on the spindle with 1000N and ..... they both deflected 0.0079mm or in stiffness terms 126N/um which is fantastic but I can't get that. So Why is it so? I built No6 base from No5 base and the bottom and corner have the same geometry. Its this corner that is the dominant (and same) deflection... Now I have the machine nailed to the earth, so I may place it on a surface, add friction and create an internal load like the tool/part would have to confirm the machine deflection would be the same.... The struggle goes on. keep Making Peter

    Attached Thumbnails Attached Thumbnails Milli a new composite mill kit-milli-no6-jpg   Milli a new composite mill kit-milli-no6-1000n-jpg   Milli a new composite mill kit-milli-no5-1000n-jpg  


  20. #40
    Member
    Join Date
    Jan 2008
    Location
    Australia
    Posts
    1523
    Downloads
    2
    Uploads
    0

    Default Re: Milli a new composite mill kit

    What about in X direction? I bet No6 is better.

    No5 column looks thicker than No6 back wall?



Page 2 of 105 FirstFirst 123451252102 ... LastLast

Tags for this Thread

Posting Permissions

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


About CNCzone.com

    We are the largest and most active discussion forum for manufacturing industry. The site is 100% free to join and use, so join today!

Follow us on


Our Brands

Milli a new composite mill kit

Milli a new composite mill kit