TaK's Sieg X3 CNC Upgrades - Page 4


Page 4 of 4 FirstFirst 1234
Results 61 to 75 of 75

Thread: TaK's Sieg X3 CNC Upgrades

  1. #61
    Gold Member
    Join Date
    Jan 2005
    Location
    Toronto, Canada
    Posts
    1695
    Downloads
    0
    Uploads
    0

    Default

    I got my sx3 down to .0007. I'm not sure if it will stay. It took a lot of careful tweaking.

    As described in Kreutz's thread, I measure backlash, with a dial indicator. I move the table forward until it pushes the indicator, back off slightly and then zero Mach and the dial indicator. Now I move the table .1" to make the dial turn exactly 1 revolution. The deviation from zero is the backlash.



  2. #62
    Member
    Join Date
    Mar 2009
    Location
    USA
    Posts
    388
    Downloads
    0
    Uploads
    0

    Default

    Just wondering cause I've not done any serious work with my SX3 yet since the conversion, but shouldn't the backlash compensation in Mach3 overcome all of this? i.e. Couldnt we just plug in the amount of backlash that you measure into the backlash comp and fix the entire problem? Seems like backlash is always going to exist, and I've seen some very nice work done with these small machines.

    Scott...

    Instructional Videos for CNC Guitar Building
    http://www.rmgvideos.com


  3. #63
    Member
    Join Date
    Dec 2010
    Location
    USA
    Posts
    1230
    Downloads
    0
    Uploads
    0

    Default

    Backlash comp is fairly violent and causes a controlled stop at direction changes leaving witness marks. getting it as low as possible decreases how violent the "catch up" is, how long the pause has to be, and the error associated with lash.

    I ran $1000 worth of parts for various customers last month (exponentially increasing so far) and every one of them had BLC turned on. I run it all the time and it does induce some error. everything from temperature to how much oil is on the screw, in the nut, on the ways etc effects lash so after making a few hundred moves one axis may be off by .003, but running the same program again a different axis could be off. I have learned to program the machine to re-ref each axis is at the end of each part.

    G53 G0 Z-.5 Y.5 X.5 (sends machine to .5" from switches)
    G91 G28.1 Z0 X0 Y0 (Ref all three axis from current position)
    G90 (switch back to absolute)
    G53 G0 X9 Y6 (rapid to convenient place to change parts)
    M30

    This has proved reliable on hundreds and hundreds of parts to within .001 ot .002 on small production. I also have leaned to rough conventional and I've started using data flute end mills (aluminum rougher finisher carbide 2 flute) and have seen metal removal close to double. Befor I was running .05-.1 DOC, now .25" DOC 60% step with a 3/8" EM in aluminum doesn't bother me at all.This is aluminum of course, AND NOT AT 2000RPM either. 4700rpm at 13ipm at those depths is pretty good for the x3 IMO.

    Now... if I try that climb cutting I have to slow it down to 8ipm to reduce vibration and the parts are .004" over/under size (outside/inside) from the lash letting the cutter "climb" out of the cut. If I had 0 lash I could rough climb and save a finish cut, cut faster, less vibration.



    All that said... for a hobbyist... who gives a crap? I focus on every cut since I'm running a LOT of parts (for this machine) and saving one minute per cycle can be equivalent (in time) to not having to run the last 10 parts. for hobbyist it is just something else to play with, a challenge to over come. Why does my 300lb friend spend thousands of dollars making his car 5lb lighter... over and over... if he skipped dinner he could save thousands and thousands of dollars LOL. No lash IS best, comp helps, but in reality very little of what most people will make need accuracy to the level we all shoot for, but why shoot for half arse?



  4. #64
    Member
    Join Date
    Mar 2009
    Location
    USA
    Posts
    388
    Downloads
    0
    Uploads
    0

    Default

    Quote Originally Posted by PriddyShiddy View Post
    Backlash comp is fairly violent and causes a controlled stop at direction changes leaving witness marks. getting it as low as possible decreases how violent the "catch up" is, how long the pause has to be, and the error associated with lash.

    I ran $1000 worth of parts for various customers last month (exponentially increasing so far) and every one of them had BLC turned on. I run it all the time and it does induce some error. everything from temperature to how much oil is on the screw, in the nut, on the ways etc effects lash so after making a few hundred moves one axis may be off by .003, but running the same program again a different axis could be off. I have learned to program the machine to re-ref each axis is at the end of each part.

    G53 G0 Z-.5 Y.5 X.5 (sends machine to .5" from switches)
    G91 G28.1 Z0 X0 Y0 (Ref all three axis from current position)
    G90 (switch back to absolute)
    G53 G0 X9 Y6 (rapid to convenient place to change parts)
    M30

    This has proved reliable on hundreds and hundreds of parts to within .001 ot .002 on small production. I also have leaned to rough conventional and I've started using data flute end mills (aluminum rougher finisher carbide 2 flute) and have seen metal removal close to double. Befor I was running .05-.1 DOC, now .25" DOC 60% step with a 3/8" EM in aluminum doesn't bother me at all.This is aluminum of course, AND NOT AT 2000RPM either. 4700rpm at 13ipm at those depths is pretty good for the x3 IMO.

    Now... if I try that climb cutting I have to slow it down to 8ipm to reduce vibration and the parts are .004" over/under size (outside/inside) from the lash letting the cutter "climb" out of the cut. If I had 0 lash I could rough climb and save a finish cut, cut faster, less vibration.



    All that said... for a hobbyist... who gives a crap? I focus on every cut since I'm running a LOT of parts (for this machine) and saving one minute per cycle can be equivalent (in time) to not having to run the last 10 parts. for hobbyist it is just something else to play with, a challenge to over come. Why does my 300lb friend spend thousands of dollars making his car 5lb lighter... over and over... if he skipped dinner he could save thousands and thousands of dollars LOL. No lash IS best, comp helps, but in reality very little of what most people will make need accuracy to the level we all shoot for, but why shoot for half arse?
    Thanks for the explanation and info on homing Priddy... Trying to zero in on the best possible backlash probably makes a lot more sense for people that have finished everything else they need to do to get the mill up and running. I have a ton of stuff left to do on mine, so even though this is super interesting, it probably falls way down the list in priorities for me.

    Scott...

    Instructional Videos for CNC Guitar Building
    http://www.rmgvideos.com


  5. #65
    Registered
    Join Date
    Apr 2005
    Location
    Canada
    Posts
    419
    Downloads
    0
    Uploads
    0

    Default

    Backlash compensation is almost always beneficial, but there is a physical limitation to how well it can perform:



    You will notice that changing direction requires that the motor instantly "takes up" the backlash in the screw by engaging the other side of the thread.

    In reality, this will always take a non-zero amount of time which means that if you are cutting a circle or whatever the machine will pause for a split second and this can mess with the finish.

    CNC Controllers also try to maintain a constant velocity when cutting (your feedrate), which becomes impossible when an axis jerks every time it changes direction.

    All together, you end up with lots of small effects that can all be traced back to backlash which affect the accuracy, finish, and speed of you mill.



  6. #66
    Registered
    Join Date
    Jul 2010
    Location
    Austin, TX
    Posts
    79
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Hello,

    I'm back to working on the CNC machine.

    Does anyone know of a good solid model for the X3? I've looked around, but come up with nothing. I've got my machine in pieces right now and I'm measuring and modeling each one but it's tedious and I'm hoping somebody's got a model.

    Right now, I have in mind to do some major changes to the ballscrews. I'll probably do just the X first and see what effect I can make.

    Did anyone take some effort to protect their screws from debris/chip with bellows or spring steel?

    Is anyone going to read these questions? This topic is probably 100 pages deep by now. Ha.

    TK



  7. #67
    Registered
    Join Date
    Jul 2010
    Location
    Austin, TX
    Posts
    79
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    I've been looking closely at the ballscrews that come with the CNCfusion premium kit. They are obviously rolled screws if you look at the landing (and the price). The X axis nut on my mill is marked: FSI1605D C0017. Unfortunately, looking at the datasheet number coding doesn't shed much light: http://takamolbearing.ir/pdf/ABBA.pdf. I'm mostly interested in the numbers for the accuracy grade, axial play and preloading, none of which seems to be reflected in the marked PN. I already knew the diameter, lead, and flange type. Ha, thanks for nothing part number.

    I also asked Michael at CNCFusion via email. He has been helpful and responsive, but only knows that the accuracy grade is C7 and thinks the axial play is 0.001" or less.

    I doubt that the axial play is 0.001" or less. The ABBA DS only specs the axial play for the non-preloaded versions of the screw and that rating is 0.0031" (suspiciously close to the backlash measurement I've made on this axis). The axial play is not rated for any screw that has a preload that I've seen. Of course I don't know what the preload rating of my screw is so its still a mystery. I thought I might be able to measure the preload torque of my ballscrew and nut, but they spin very freely (no preload?), but most preload ratings seem to be < 4 oz-in anyway so it would be tough to detect a such a small one.

    I've compared these numbers to some other ballscrew vendors. Here are some other axial play numbers for comparison including some very expensive screws:

    rolled C10: 0.004"
    rolled C7: 0.0011"
    ground C7: 0.0011"
    ground C5: 0.0002"
    ground C3: 0 (preload 1.5 - 6.0 N-cm)

    Another potential source of axial play is, of course, the fixed bearing block for the axis. The kit uses back to back angular contact bearings which is good. My screw doesn't spin very freely in them anymore, I think there are chips stuck in the bearings. These would definitely benefit from having some sealed bearings. Anyway, I tried to remove the nut on the ballscrew to see the markings on the bearing and clean the debris out, but I haven't been able to hold the ballscrew tight enough to remove the nut. There are no wrench flats or similar to hold the screw while you work the nut... it's annoying. The bearings are press fit into the aluminum bearing block. I'm assuming this is quite tight and there isn't play in that fit.

    Lastly, I'm pretty sure by this point that the nut block is well mounted to the saddle. It would be very nice if the locating pins were being used and not just the two screws, but I am fairly confident is isn't moving around. I once let a lot of red Locktite leak out of the screw holes to effectively glue the block in place and adjusted the gibs to a light pressure and it made no difference on my backlash reading as best I could tell.

    I'm still trying to think of a good way to protect the screws from getting debris on them. I see that there are chips and things on my X axis screw which I thought would have been well protected. I wouldn't want to invest in expensive screws unless I'm confident I won't wreck them with chips.

    I found the very well-made solid model of the SX3 that Hamzter put together. I guess it has been quite a while since I've been on these forums... Anyway, I'm working on adapting that for the X3 and checking some of the critical dimensions and fixing the Solidworks import issues. The features imported mostly correct, but there are a bunch of random feature errors like planes defined to be fileted and they don't intersect...



  8. #68
    Registered
    Join Date
    Jul 2010
    Location
    Austin, TX
    Posts
    79
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    I spent time working through Hamzter's SX3 model in Solidworks. All the mates had to be redone due to the importing. I need to do the modeling for the X3 head, but for the mean time I've left the SX3 head off as well as all the parts removed from the CNC conversion. I went and modeled the X-axis components of the CNCFusion kit.





    The hole alignment on the saddle is good and on the stepper sides it's OK, but not perfect. There is an issue with the hole alignment on the fixed support side.




    Here is the issue with the fixed support side:



    The asymmetry is on purpose, but there is an offset to the slots I measured on the kit and what's in Hamzter's model. I'll do some measuring and figure out which of us made a mistake. It could also be that there is a difference in our mills, but that isn't super likely since the other parts match up.

    There is very little clearance for the flange of the ballnut; only around 1mm on the top and bottom. That's pretty tight. The same goes for the bottom of the fixed side plate assuming you want to over-travel the edge of the ways.

    Attached Thumbnails Attached Thumbnails TaK's Sieg X3 CNC Upgrades-tak_cncz_001-jpg  


  9. #69
    Registered
    Join Date
    Jul 2010
    Location
    Austin, TX
    Posts
    79
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    I spent time carefully measuring the worktable and found the problem. The worktable is modeled as symmetric from the left or right views. You can see this in the last picture of the previous post. In reality, the worktable is not symmetric and the ways toward the front of the worktable are longer than the ones toward the back. I don't know if this is an issue with the original model or just the import.

    Now that I have the worktable correct, I see that there is an issue with the saddle. I'll have to get this correct also before I start cutting anything.

    I began importing some possible components for an X-axis upgrade.

    TK



  10. #70
    Registered tauntdesigns's Avatar
    Join Date
    Nov 2005
    Location
    USA
    Posts
    655
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Nice model.......... I wish I had done a solid model when I had mine apart years ago

    JK

    Walking is highly over-rated


  11. #71
    Member diyengineer's Avatar
    Join Date
    Feb 2010
    Location
    USA-WA STATE
    Posts
    3447
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Hows the cooling system working? Just bought one myself so far havent tried it out. I mounted it about 2' below where the dispenser is, so i hope i wont' have issues. Someone else said they did the same without problems, but the manufacturer reccomends it be at the same height.



  12. #72
    Member lancut's Avatar
    Join Date
    Nov 2008
    Location
    USA
    Posts
    412
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Nice! Are you planning on posting the model for download. Thanks.

    My 2¢


  13. #73
    Registered
    Join Date
    Jul 2010
    Location
    Austin, TX
    Posts
    79
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Hello,

    My progress has been pretty slow due to limited time to work, but I'll share some stuff:

    Hows the cooling system working? Just bought one myself so far havent tried it out. I mounted it about 2' below where the dispenser is, so i hope i wont' have issues. Someone else said they did the same without problems, but the manufacturer reccomends it be at the same height.
    The Fog Buster works fine, but I've mostly been cutting AL which doesn't really need coolant. I had expected to do more work in steel which is why I bought it, but that hasn't happened so far. I usually adjust the position each time I change the tool; I haven't really settled on a fixed position.

    Nice! Are you planning on posting the model for download. Thanks.
    Um. Maybe. You can definitely go download Hamzter's model right now from his thread. As for the one I've been modifying/fixing, I might post it. The problem is I've only worked on a couple parts of it and there are probably lots of other things that need work. I don't want to keep posting partial work and I don't really have a timeline for being 'done' with it.

    I've determined it's very challenging to measure the center locations of tapped holes. After measuring and remeasuring about 4 times, I think I have enough accuracy in the model that I can use it to make engineering decisions. I made the saddle over again from scratch and I modified the worktable main shape to not be symmetric as it was originally drawn. You can see the difference in the worktable by the side views from the original.

    I've only really spent time on components that effect the X-axis like the worktable and saddle so far. I'm sure there's work to be done on the rest at some point also.

    Here's how the mounting alignment looks:






    Notice that there is still some misalignment of the holes to the mounting plates of the CNCFusion kit. I believe this is real because I have to rotate the plates a small amount on my machine to get them to align and not put pressure on the screw.

    I'm done picking parts for an upgrade of this axis, but I still have to finalize on the shape of the new mounting plates and then make a final decision on whether or not to spend the money. I'll post some screen shots of the solid model of the parts I'm thinking about next time.

    TK

    Last edited by TaK; 04-10-2014 at 02:45 PM.


  14. #74
    Registered
    Join Date
    Jul 2010
    Location
    Austin, TX
    Posts
    79
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Hello,

    So here's what I came up with for the x-axis:







    A few points for thought:

    - I wanted each end plate to be clean, fit in a comfortable work area when I'm cutting it, and maximize contact area with the end of the worktable to minimize tilting under load
    - I'm looking at a component fixed-side back-to-back angular contact bearing housing; it keeps the bearings enclosed to keep chips out and I never liked the press fit arrangement in the CNCFusion kit
    - the ballnut has a larger flange than the CNCFusion kit ballnut, but it should just barely fit under the saddle
    - for the screws that hold each end plate, I'm looking at a round through hole with about 2mm clearance all around; I'll have to use a quality washer, but I've chewed up the CNCFusion kit screw slots because they are too thin and I didn't want to do the same when I really tighten the screws down

    It allows each side to over-travel some distance:




    The over travel on the stepper side is limited by the stepper motor frame itself. The CNCFusion kit did not allow over-travel on the stepper side. The over-travel on the fixed side is limited by the size of the bearing housing.

    Now I just have to decide if the cost is worth while. Ha. Maybe I'll cut the plates while I think.

    TK



  15. #75
    Member X3Miller's Avatar
    Join Date
    Sep 2021
    Posts
    2
    Downloads
    0
    Uploads
    0

    Default Re: TaK's Sieg X3 CNC Upgrades

    Hi..
    What kind of motor are you using and what else have you done to get hi RPM.?



Page 4 of 4 FirstFirst 1234

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

TaK's Sieg X3 CNC Upgrades

TaK's Sieg X3 CNC Upgrades