Shop OBEX P1 Docs P2 Docs Learn Events
Linear drive Screw questions. - Page 2 — Parallax Forums

Linear drive Screw questions.

2»

Comments

  • CannibalRoboticsCannibalRobotics Posts: 535
    edited 2009-03-22 17:52
    Oh yeah, on the drives, all you need are four beefy NPN or darlington transistors, a high watage 10 ohm resistor and a power supply. There are dozens of websites with cheapo designs for stepper motor controllers.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Signature space for rent!
    Send $1 to CannibalRobotics.com.
  • chaosgkchaosgk Posts: 322
    edited 2009-03-23 02:41
    Well, I installed my first drive screw tonight just for testing the X axis with just a scrap piece of HDPE I had molded for a drive nut. I'm currently using my dewalt 18v drill to run the screw back and forth. Wow, it's amazing how something as simple as a test screw mounted in the system and seeing it go back and forth under power can really make a guy feel good. I had my drill on the lowest clutch setting it had and it still moved it back and forth. I had to go to about setting 5 (out of 20) if I put a little resistance on the gantry while it was moving, but it did have plenty of power at that setting.
    I even had my daughter stand with her feet behind the gantry as it moved back and it pushed her across the floor with no problem so I'm not a bit worried about my steppers having enough power to do what I need it to.
    I am a little concerned with the speed of it. I'm not sure how fast CNC's are supposed to operate, but using the drill on it's highest setting, it takes about 30 seconds to travel the full 30".
    I may have to see about getting some ACME screws and nuts if these all thread rods don't work out too well. It's modular enough to be able to switch all of those out in a night.
    Now I just need to work out a way to attach the motors and get to finding a driver.
    Is there a way to directly control a stepper off a parallel port? I'm not looking for high speed or anything, just a way to see if these steppers move.
  • chaosgkchaosgk Posts: 322
    edited 2009-04-02 02:24
    Ok, new question for you all.·· I talked about running two steppers in parallel off the same control channel, we decided that wouldn't work.· I just ordered my stepper controller, it is 4 channel, would it be feasable to use a stepper on each side of the Z rails·to lift the entire assembly using two of the channels on the controller.· My motors don't have an encoder in them, I was thinking of using home switches on each side and have it setup so that every so often the Z axis goes home to make sure they are in sync and then continue with it's job.· On normal running for signs and drilling, it is only going to be moving up and down 1/4"-1/2".· I have done some testing and just turning the motors by hand and using a digital caliper, it looks to have a resolution of about 1/1000th of an inch, and I coudn't detect any backlash with the screws and threaded plastic i'm using.· I have the X and Y motors mounted and aligned and will be mounting the work surface tonight.· Just waiting to see what the best option is for the Z axis.· If the two on the ends don't work, I'll build the rail and mount for the Z axis to attach to the Y slide bearing.
    ·
  • Carl HayesCarl Hayes Posts: 841
    edited 2009-04-02 02:50
    chaosgk said...
    ·would it be [noparse][[/noparse]feasible] to use a stepper on each side of the Z rails·to lift the entire assembly using two of the channels on the controller
    Sure, why not?· You could even make your own sync-checker with a couple pieces of sensing tape (like the tape used at the ends of recording tapes) on the shafts,·and a pair of contact fingers to sense them.· If the two shafts don't agree, it's time for a calibration cycle.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · -- Carl, nn5i@arrl.net
  • chaosgkchaosgk Posts: 322
    edited 2009-04-08 05:28
    I decided to go with a single servo on it's own Z axis rail, after getting the bearing block adjusted correctly it works great. My stepper drivers came in today and wouldn't you know it, I don't have a single function computer with parallel ports (I could have taken my daughters but she would have been mad). My old laptop has one, but Windows 98 so the software wouldn't work. After getting a 98 version, it installed a update and it wouldn't boot so now it's being upgraded to xp. The Z axis is ready for the motor to be mounted and attached to the Y axis bearing slide. I may in fact take the day off from my real job to work on it.
  • JasonDorieJasonDorie Posts: 1,930
    edited 2009-04-09 19:53
    chaosgk said...
    Would it be feasable to use a stepper on each side of the Z rails·to lift the entire assembly using two of the channels on the controller?
    Mine uses a single stepper for each axis.· The main axis uses two drive threads to keep it square, but they're both connected to a single stepper with a timing belt:

    http://www.facebook.com/photo.php?pid=2270211&l=b278f9c355&id=671403972

    The gray thing is an electrical conduit coupler mounted on bearings to act as a belt tensioner.· The setup·works quite well.· You get 2x the friction losses, but it hasn't been a problem for me.

    Jason
    ·
  • willy1067willy1067 Posts: 107
    edited 2009-04-10 04:48
    Cool stuff, I put a router and laser on my CNC table

    ·http://forums.parallax.com/showthread.php?p=656332

    http://www.youtube.com/watch?v=LGQ7sTf4diM· video CNC routing

    for software, get KCam4, is free for 90 days, so you can twick your motors

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Fernando Gomez


    ·http://pinellas-sign-manufacturer.com

    Never compare yourself with anyone else, there will always be someone bigger·or·smaller·than you.

    Post Edited (willy1067) : 4/10/2009 5:00:17 AM GMT
  • chaosgkchaosgk Posts: 322
    edited 2009-04-12 06:29
    Wow, awesome job on that. Mine is functional finally, have been doing some test cuts and such. I'm having trouble figuring out what software to actually design what I want cut in. I'm using a open source program called inkscape and it is working ok at best. I have Photoshop CS 9, and a few other different graphics programs, but I can't figure out how to tell the software the diameter of the router bits or any specs like that. How to tell the system that it is supposed to follow the outside of the letters it is cutting only and not try to fill them in. Or if it is going to fill them in, do it correctly.
    I am using lazy cam as the converter to tell it the depth of cut, size of the bit and cut speeds but I am getting crappy quality on the letters.

    The CNC itself talks with Mach3 just fine and is accurate enough for what I am doing with it.
    If anyone has any suggestions for a better interface for going from graphic to gcode, please let me know
  • JasonDorieJasonDorie Posts: 1,930
    edited 2009-04-13 07:13
    You need a CAM program to generate the tool paths for you. Check out CamBam or MeshCam. I bought MeshCam and have been quite happy with it.

    Jason
  • chaosgkchaosgk Posts: 322
    edited 2009-04-17 20:22
    Ok everyone, riddle me this because i'm very confused.
    On my CNC drive screws, I was using the 5/16th" x 18tpi threaded rod for the linear drive. I had trouble with the X axis drive nut because I didn't have my bearings tightened on the blocks so it wore down the nut pretty bad because of the extra play.
    Anyway, since I had it apart, I wanted to try and get more speed out of it so I switched to 1/2" x 13tpi rod. I turned the ends down to the original 5/16th" for the bearings and 1/4" for the connection to the motor.
    I made the steps per inch adjustments in the software and ran some test runs with the rapid rate. With the smaller rod, I was limited to around 20 ipm transit time. With the new rod, I was pushing it to 75 ipm before my rod started to wobble in the middle because I didn't have it locked into the bearings and the drive nut wasn't attached yet to keep the wobble out of it.
    How can I get that much more speed out of it. I went from probably 360 rpm to close to 1000rpm. The motor by itself was having trouble going over 360rpm with nothing attached, it would buzz and slow down.
    What is the reasoning behind being able to push 3 times the speed to the screw itself and in turn getting 4 times the transit speed from wha I thought possible for these little motors.
    I am pushing the same voltage to motors, between 28 and 30. If I drop under 28 it has trouble getting up to speed and buzzes a lot.

    Why, why, why? I'm so confused.
  • chaosgkchaosgk Posts: 322
    edited 2009-04-20 14:53
    To answer my own question, I found out with some testing and research from my brother that with stepper motors, they actually like to have some load on them. It turns out that the larger rod was providing some rotational inertia and acting as a basic flywheel.
    I added a 2.5" round steel plate to the end of one of the smaller screws and was able to push it up to 40ipm. I'll be adding more weight to be able to increase this even higher soon.
  • JasonDorieJasonDorie Posts: 1,930
    edited 2009-04-20 18:13
    Chaos - Check out this page:

    http://solsylva.com/cnc/dampers.shtml

    Following this tip, I added a couple $4 rubber wheels to the rear shaft on my steppers and was able to take my rapids from 50ipm to about 185ipm.· It's possible that they'll go even faster, but I have to increase the 'tick rate' of the OS driver to test it.

    Jason
    ·
  • chaosgkchaosgk Posts: 322
    edited 2009-04-21 13:45
    Thanks Jason, that was a great help. I was looking for a good source of a 2 1/2" fly wheel and didn't even think of casters. I think I actually have some of those at home already. I'm guessing that the spring is in there to provide some brake to the wheel so it can flex on acceleration and deceleration?
    I may have to add this to the inside of where my drive screw runs instead of on the end as I only gave my shafts about 1/2" of extra lenght and the stepper doesnt' have any extra on the back.

    What are you using for 3d software? I did some looking at the Vectric software and am amazed at how well it works, however you cannot actually save anything to be used by the cnc and the prices are a little heavy for me.
  • chaosgkchaosgk Posts: 322
    edited 2009-04-22 16:24
    I used a mod based on the link you sent for the dampers. Instead of using the 2 1/2" rubber wheels, I went with 2" metal casters with a spring tensioner to adjust them. I went from 13ipm on my Z axis to about 30 which made a huge difference. I also replaced my metal cutout flywheel on my Y axis screw with the same metal wheel and took it from 40IPM to 60IPM with plenty of torque to spare.
    I'm having trouble with my X axis shaking a little, anyone have a simple solution for adding an encoder to a stepper that doesnt' have one? I'd like to modify my X axis to have two steppers, one on each rail. I have a spare channel on my controllers and the Mach 3 software will let me slave to another master channel so they run in sync.
    Any problems with this? If that doensn't work, I'll have to use the cable solution to get them to run parallel.
  • JasonDorieJasonDorie Posts: 1,930
    edited 2009-04-23 07:04
    For 3d software I'm using Alibre designer, though I'm going to look at Rhino too - I've heard decent things about it. I'm still using the free versions for now. They do tend to me pricey.

    Jason

    Post Edited (JasonDorie) : 4/23/2009 7:10:44 AM GMT
  • chaosgkchaosgk Posts: 322
    edited 2009-04-29 15:13
    Well, I fried two control chips last night.· I'm not exactly sure how it happened. I am switching to 1/2" acme 10tpi screws so I tried switching the controller to 1/2 stepping to keep my resolution high.· When I powered the system back up, I saw one of the chips release some of it's magic smoke and I powered it down immediately.· I switched it back to full step and powered it up and the other chip released it's magic smoke as well. I was running at 30v at the time like they had been running at for the last few weeks without problems. Anyone have a guess as to why both chips on this board fried, but the other board is fine?
    ·
Sign In or Register to comment.