1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. PLEASE: Read the FORUM RULES BEFORE registering!

    Dismiss Notice

Grizzly G0704 Cnc Conversion

Discussion in 'MACHINE BUILD LOGS' started by cozmogeek, Jan 9, 2016.

  1. jbolt

    jbolt United States Active User H-M Supporter-Premium

    Likes Received:
    757
    Trophy Points:
    93
    City:
    Mountain View
    State:
    California

    -Return to Top-

  2. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    That's what I was thinking of. It's basically what Hoss does on his DVD, except his is 3D printed, too. A tube with and ID that will pass the machined end of the ballscrew and slide over its end, but has an OD smaller than the minor diameter of the screw. Then the ballnut is unscrewed onto the holder, trapping the ball bearings.

    nut_tool.jpg
    I don't have any stock handy to do this, but hadn't considered taking the whole thing up to the Borg and seeing if a piece of PVC will work.
     
  3. coherent

    coherent United States Active Member Active Member

    Likes Received:
    26
    Trophy Points:
    18
    City:
    Flagstaff
    State:
    Arizona

    -Return to Top-

    FWIW I opened the file for the ballnut removal tool (tube) and it's .53in or 13.46mm. My guess is a piece of 1/4" pvc sched 40 would prob work. They show it has an outside diameter of .54in with tolerance +/- .01"

    Even a piece of hose or vinyl tubing with a 1/2" outside diameter should work... especially with a little duct tape if you need to enlarge the diameter a little. Just guessing you don't have a lathe or this prob would be a issue at all.
     
  4. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    No, I have a lathe. I just don't have any stock of that size to turn. I'd have to order something. There's no place locally (that I know of) to go get some plastic stock to turn, unless it's something the big box home centers carry.
     
  5. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    An intermediate step along the way. Y-axis ballnut.
    Y-BN_Trimmed.JPG
    Used a diamond grinder. BTW, these are the linearmotion ballscrews from Chai in China, and this is good steel. It's not case hardened; it doesn't get softer after you take a hard skin off. It's hard this whole depth.
     
    JimDawson likes this.
  6. JimDawson

    JimDawson Global Moderator Staff Member Director

    Likes Received:
    3,566
    Trophy Points:
    113
    City:
    Sandy
    State:
    Oregon

    -Return to Top-

    Probably not the best for grinding steel, even hardened steel, they tend to load up quickly and then just make things hot. Diamond is great for grinding rock, glass, and ceramics. The better choice for steel is just a standard 36 grit grinding disk.
     
  7. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    I left out a word. I used a wet grinder that was designed for grinding rocks. The water flood washes the grinding dust out constantly. No one in their right mind would go buy one for this use, but since I have one, it was convenient and worked well.

    By the way, anyone thinking of the mod and wondering, the width across the widest points on the ballnut was about 1.88". The width of the area it has to clear in the G0704 base was 1.84. That sounds like just taking a total of .050 off the piece would work - 25 each side. I made mine look like the one in Hoss' pictures and it came out 1.55. Way more than necessary.
     
    JimDawson likes this.
  8. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Just a post to keep this thread alive. I'm still working. Latest is finishing the chip tray, the enlarged one that's not on the DVD, but buried in his videos on YouTube and his CNCZone forum.
    Tray&Mill.JPG
    This is the tray, rev.1, which has already been modified. See the big hole at the top of the middle panel? That's the way it appears on the jpeg drawing Hoss posted on CNCZone. Only the videos move it forward, closer to the front edge, and enlarge it from 1 1/2" to 3" for a common sink drain. I got that done after this picture was taken. It's 64" tall in this pic, which is left to right in use, and 36" front to back.
     
  9. coherent

    coherent United States Active Member Active Member

    Likes Received:
    26
    Trophy Points:
    18
    City:
    Flagstaff
    State:
    Arizona

    -Return to Top-

    Nice job on the chip tray. I wish I had room for something that big, but as an alternative I'm planning a smaller partial (1/4 to 1/2 height) enclosure. I plan on getting a new tig welder soon so thought I'd make it my first tig project and use aluminum. Be sure and provide some photos when you get yours sealed and mounted!
     
  10. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    If you have Hoss' DVD, he has a smaller chip tray in the "Projects" folder. It's shorter front to back, and the Y axis motor sticks out front, and 59" left to right while the new one (like mine) is 5" longer. 5'4". The left to right size is pretty much fixed by the length of the table with the X-axis motor and Hoss' mount.

    Hope that's useful.
     
  11. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    It has been a couple of weeks since my last post, so just a little update. I'm rearranging the shop and getting the pieces in place to start taking the Grizzly apart. The big change is in this view of my shop a couple of days ago.
    NewShopLayout.JPG
    You can see the mill is on a rolling base, the one Grizzly sells for it. You'll note the new wood chip tray is over against the left wall. While I was figuring out how to lift the 350lb lathe and base cabinets with the shop crane (visible at the left), my wife got behind it and found she could push it on the smooth floor. My wife is rather petite: 4'10" and a small fraction of the weight of the lathe on its stand. Between the two of us we pretty effortlessly rotated the lathe 90 degrees to the direction it was in, and slid it toward the right - toward the camera in this view. Guess that means I don't really need the mobile stand I bought for the lathe but hadn't put together.

    My big remaining jobs are:
    • Fix/build/buy a computer to control this CNC mill. I'm almost definitely switching over to LinuxCNC for this from Mach3, which says the computer will be Linux instead of another XP box like I have. (I use Mach3 to run my Sherline CNC mill and lathe). Because of past Linux experiences, this is a scary step to me. I've run Linux a few times before, and it always seems if you want a computer to just do common tasks, it's great, but if you need anything unusual it can be a total disaster. Unusual things like a Linux driver for a USB-connected motion control board. The computer needs to be a priority now, because I'm going to need the CNC controller to control the next part:
    • Build the box to hold the hardware for motion control. I think I have all this hardware, mainly the stepper motors and their controllers from Automation Technologies. I have an emergency stop switch, limit switches, connectors for cables, and all those niceties. The limit switches get built into the next part:
    • The CNC conversion to the mill. I'll need to take the mill apart to the subassembly level for this - maybe not to individual pieces of hardware. I'll need to modify the base of the machine with a drill and angle grinder. This is where the adventure really starts, because if I did something wrong and need the mill to fix it, I'm SOL.
    • While the system is apart, I need to add the oiling system, which will pump light oil onto the leadscrews and a few other places. I have the oil fittings, but not everything, just most.
    • When the system is back together and cutting metal, that's the time to build the enclosure and cooling fluid system. I have most of those parts, too.
    Whew. I wonder if I can get it done before my one year anniversary of retirement in December?
     
  12. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    The best laid plans of mice and men...

    I did complete the first two items on that list in my last post. I got LinuxCNC running on the designated computer. I got my motors running in a temporary setup. It's just that between the two computers, neither one is really good. They both have odd intermittent problems. The Linux machine won't start most of the time, and has to be turned off and on to get it to stay on. When I eventually get it running, it seems to stay running. The Windows/Mach3 machine will start and run, but then freezes sometimes when I move the mouse. Maybe once every couple of months. It has never hung up while running CNC, so that's a good thing. I can't tell if the system would halt or keep driving until the system hits the limits, breaks tools, ruins work pieces, all that.

    Right now, I'm planning to use the Mach3 machine with two configuration files: one for the old Sherline systems, one for the new Grizzly. Meanwhile, I'm trying to make sure everything will fit in the box I've got dedicated for it by modeling it in Rhino, my 3D modeling program. The model includes a fourth motor controller that I don't have, for an eventual rotary table.
    ControllerBox4.jpg

    Still plugging along.
     
    TomS and JimDawson like this.
  13. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Since it has been about three weeks since the last one, time for an update.

    The last picture showed my model, at that time, of the test box. Things continued to move around in the box and on the rear panel, and other things needed to be ordered and received. The latest model, which I'm calling the "final version" (hah!) is this:
    Box_Final_Model.jpg
    Sharp-eyed readers will notice that the breakout board has moved toward the viewer and the chassis is now cutout with a sloppy tolerance opening for the DB-25 hard mounted on the BoB. I simply didn't have a simple, clean way of connecting it where it was. The fan is obvious, but in the lower left is a new part I found. This combines all of the AC input parts into one much smaller part: the AC socket, the power switch, a fuse holder and an EMI filter, all of which were separate parts in the previous version. The four DB-9 connectors move from that group of four on the right into a single row across the top of the rear panel.

    Once I had everything located, it was time to locate all the holes. The bottom was the simplest piece, so the first one I attacked. This is the bottom panel put into the rest of the chassis as a test to make sure everything fits as intended. It all did.
    ChassisMockup.JPG
    Once the bottom was drilled out and tapped in the right places, it was time to go over to the back panel. It has four DB-9 connectors on the top. I had a file I found online somewhere that would cut them, but ran into some problems with it cutting too big. I ended up writing my own file to cut them, but wasted a few days getting through that. After two test cuts, I was ready to move over to the real thing.
    BackPanel.jpg
    That's my progress as of Monday. I have two rectangular cutouts to make, and then drill out the fan holes; right now I'm planning to change from the 1/4" holes over to a #5 center drill, so that I can potentially cut and debur in one step.

    I wasn't sure it could, but thankfully my A2ZCNC/Sherline mill can handle the entire 16 x 6 1/2 panel to cut some of these features. Especially if I do a little at a time.

    Still plugging along. I always do my best to answer questions, but I'm in the highest danger zone for Hurricane Matthew. I expect to be without power for a while over the next few (???) days.
     
    TomS likes this.
  14. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Time for another update. All of the metal for the test box has been completed and I've started wiring things up for good. The back panel required three separate setups, which is painful, but it's actually larger than my Sherline/A2ZCNC hybrid can handle. The array of holes is for the fan; the two rectangular cutouts are for the Breakout Board (the longer, narrower one on the left) and the AC Interface (Corcom line filter/switch/fuse/AC receptacle). The panel is upside down in this view; the row of DB-9 cutouts goes at the top. The BoB is on a mount: a piece of 4x6 alumimum sheet, 1/8" thick. It has six holes in it, but I did it on the mill as a precision drill press. It screws to a small piece of 3/8 aluminum that's tapped 6-32 in four places: two to mount to bottom plate and two to mount to the 4x6x1/8" sheet.

    DrilledPanel.JPG

    Several test fits were checked. I had to make a minor correction to the BoB mounting bracket, moving it back about 1/8" from the rear panel. I laid it out by hand and messed up.
    TestFit1.JPG
    Note the BoB isn't really mounted, just hanging there on some screws.
    TestFit2.JPG
    This a test fit of the Corcom part.
    Finally, an inside view of everything.
    TestFit3.JPG
    Again, the BoB and mount, the fan, and the AC input aren't mounted "for real". No hardware.

    The last piece to machine was the front panel, which gets a mess of holes (for air flow) and a large hole (7/8") for the Big Red Stop Right F***in' Now switch.
    FrontPanel.JPG

    Like I say, I've only just started wiring, so it will take a couple of days to hook everything up, since I'm trying to be neat. I had the demo setup on my benchtop in an hour or so, but I had no cares about making it look nice. Hopefully, within a week, I'll be able to post a video of the computer running motors on the benchtop.
     
  15. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    You know I looked at that back panel for two days before I realized, "hey!... that switch is in the wrong corner!" It was supposed to be in the upper left corner. See, I'm used to looking at it from behind, from the inside looking forward (see the models a few posts farther up). It's in the upper right viewed from behind.

    That said, it's not a problem: everything fits, and the switch works. It's just not the way I drew it in CAD, so it's a personal pride thing. The box is wired now. I don't have the cables to connect it to the motors made yet, though.
     
    JimDawson likes this.
  16. Boswell

    Boswell United States Hobby Machinist since 2010 H-M Supporter-Premium

    Likes Received:
    87
    Trophy Points:
    28
    City:
    Cedar Park
    State:
    Texas

    -Return to Top-

    Switch on the left or the right, it still looks Fantastic. Great work.
     
  17. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    It works! Got the box finished, completely wired. Made all the cables and interconnects. Plugged in the three motors and everything works on command.

    This is the box on my messy bench when I finished it.
    TestBoxBench.JPG
    A closer look inside
    TestBoxBuilt.JPG
    It's all ready for adding a fourth axis. I don't have a rotary table, yet, but I went ahead and ordered another motor and controller so I can add that some day and all the hardware will be compatible.

    Here's a video of the motors turning for a few seconds. Left to right: X, Y, Z. The very end of the video shows me pushing the Big Red Stop Right F***in' Now button, then showing the motors not moving, and finally doing a clumsy camera pan over to the computer to show Mach3 responding with "Hey, you told me to stop".


    After a little minor cleanup, it's time to take the Grizzly apart and start doing the actual CNC conversion.
     
    JimDawson, jbolt and TomS like this.
  18. cs900

    cs900 United States maker of chips Active Member

    Likes Received:
    148
    Trophy Points:
    43
    City:
    binghamton
    State:
    New York

    -Return to Top-

    great work! It's always very rewarding when things work the first time.
     
  19. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Just a photo of the controller box with the cover I made for it. It came with a black sheet steel cover like the bottom cover most everything is mounted to, but my wife remembered a piece of plexiglass she has been moving around the house since the early 80s. She thought it would be a good top for the box, since there are diagnostic LEDs inside and a see-through cover would make it easy to check those lights if something starts going wrong. Seemed like a good, practical idea to me. Finished, it looks like this:
    ControlBoxCovered.JPG
    Yeah, it was orange plexiglass. I'm sure I'm the only one on the block who has one. Or the only one in the world. For some reason, I decided to use the junk screws from my powder coating adventures I talked about, so those are blue. Gives it kind of a wild color scheme.

    The really observant will notice there are four motor controllers in here, while the picture a week ago had three. The fourth axis is installed and tested. I don't have a motor just yet, but that's OK. I don't have anything for the motor to connect to!

    Does anyone know of a good rotary table for a mill the size of a G0704?
     
    JimDawson likes this.
  20. Boswell

    Boswell United States Hobby Machinist since 2010 H-M Supporter-Premium

    Likes Received:
    87
    Trophy Points:
    28
    City:
    Cedar Park
    State:
    Texas

    -Return to Top-


    GROOVY MAN.
     
  21. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Well, time for an update, and it's time to do being this:

    Open.JPG

    Taking the machine apart to put in the ballscrews and all the hardware I've built. Took the table off it a few days ago. Yesterday and today I used my Sherline CNC mill to cut out the pocket for the X-axis ballscrew. This is a test fit.

    X-axis-TestFit.JPG

    The table weighs 10 pounds and it's a bit large for a Sherline, but nowhere near as large as the front and real panels for my motor driver box. I've never cut cast iron before and it's nasty, messy stuff. The chips are about the size of 100 grit abrasives, and are magnetic. They'll cling to the cutter and mess the cut up. I used Tap Magic and the bit never felt warm to the touch, even though the oil was smoking. Between using the Tap Magic and running the shop vac, I could keep the magnetic swarf from messing me up too much.

    The other thing I ended up doing for much of the last week was getting moved to a new computer and getting up and running in LinuxCNC. My Mach3 computer was getting flaky and that's a little scary when you're moving the table and workpiece around a full HP motor driving a cutter.

    Next is to take the head stock off and then the Z-axis. Cut the oil channels in this piece. Add the oiling system. Put all the new hardware in place...
     
    JimDawson likes this.
  22. TomS

    TomS Active User H-M Supporter-Premium

    Likes Received:
    401
    Trophy Points:
    83
    City:
    Redding
    State:
    California

    -Return to Top-

    You're moving right along. Looks good.

    Tom S.
     
  23. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Since the last update, I've run into a difficulty. I was preparing to cut the oil grooves on the Grizzly's table with my Sherline/A2ZCNC system and was finding zeroes and setting up to start cutting. I sent the table to an X-axis value where the cutter should have been left of the table by a quarter inch or so, but it wasn't. That made me suspicious and I started to make the table move left and right. It started losing motion on its X-axis. Not losing steps, losing motion. I could see the motor turning while the leadscrew wasn't turning. Sometimes. It was intermittent and irregular, like something accidentally slipping. At one point, I wrote a little "torture test" file that just moved the table back and forth 8 inches. After about half an hour, the left end of the motion had drifted right half an inch. That would have made a real mess out of the oiling grooves.

    Since I had a motor coupler break before (long enough ago that I don't remember any details) that's what I thought it was, and I started taking things apart to get the coupler out. Sherline uses a funky system in their motor mounts. Still, it may be an odd system, but the Y and Z axes use the same system and are both fine. The leadscrew for the X-axis ends in a small (half inch long?) portion that's threaded 1/4-20 and ends in a small, conical taper that engages the matching taper on the coupler. The coupler is the part that transfers the motor's motion to the leadscrew. The coupler is attached to the lead screw with a 5-40 screw down the axis of the leadscrew, and a 1/4-20 nut pulls the coupler tight against ball bearings on one side and pulls itself tight against the ball bearings on the other side. I think of the two attachments as squeezing the coupler between the 1/4-20 nut on one side and the 5-40 screw pulling the leadscrew taper tight against the coupler. Personally, I'd like a Woodruff key or something like it, but there's not enough room in their mount for that!

    A little diagram:
    SherlineMM.jpg

    I removed the stepper motor, then left the Allen wrench in the hole where the setscrew on the motor shaft is, so that I could remove the 5-40 screw and see if the coupler was broken. While holding the coupler motionless, I could turn the leadscrew and move the table by turning that 5-40 SHCS.

    It took a bit of struggle but I was eventually able to pull that screw out of the end. It felt like it was stuck in taffy, very hard to remove. When I got it out, I could see the threads were damaged.
    5-40-SHCS-X-Axis.JPG
    The threads on the 1/4-20 also looked squared like that. I was able to run a die over the 1/4-20 threads and clean them up, but I don't have a 5-40 die. I just bought a 5-40 tap to clean up the inside of the leadscrew and a few replacement screws. I had two replacement screws and put one into the leadscrew. It was very hard to thread in and remove; when it was removed it looked like the original, damaged screw.

    I put this system together back in February of '07, so almost 10 years ago. It has gotten fairly light use, and gone months without use, but it has been running almost 10 year. It doesn't seem like these screws would have gotten squared off over time. I mean if either one had backed out, the system would have stopped moving then. Which means I must have done this when I installed the A2ZCNC upgrade. Still, it ran for a long time and only now started slipping. Since there's really only a few parts here: the coupler, 1/4" nut and 5-40 screw, I ordered all of them and I'm going to rebuild the axis as new. A2Z says "Install the motor mount and coupler on the leadscrew. We recommend a slight amount of Loctite red on the taper portion of the screw. Be VERY careful to not get any into the 5-40 hole in the screw or in the bearing. Snug the 5-40 screw, but be careful to not break it by overtightening." Before everything goes back together, I'm going to degrease the inside of the leadscrew, make sure everything is "squeaky clean" before I put it together, but I think I'm going to use blue Loctite.

    While waiting on the parts to rebuild the axis, I went to make the ballnut removal tool that we all talked about a while back (August). I bought a 6" long Delrin rod to use. It turns out that the dimensions of that piece, which is a 0.400" ID over the full length, aren't right. The Y-axis ballscrew I have has a length of 0.5" OD for about about 1-1/4".
    Tool.jpg
    I had a bit of a problem making this on the big lathe. With a wall that's about .015 or .020 thickness, I had to turn with a live center, and at one point I pushed it too far into the plastic, causing it to buckle just a little. When the cutter went over that spot, it not only got too thin, it tore in one place.
    Ballscrew-Tool.JPG
    It turns out the required OD is almost exactly right for a piece of 1/4" nominal schedule 40 pipe. I may get a piece of black steel pipe and bore the inside diameters to the right dimensions. Or just try using this. It's a convenience; a one-time tool. It's a PITA if I have to replace all the balls, but shouldn't be a big deal. Assuming the balls don't roll all around the shop and into invisible corners.

    I'm really hoping to get this project done by the end of the year, but I have no idea how long it's going to take to get done. I keep telling myself my hobby isn't building CNC machines, it's making stuff. I want to get back to making stuff.
     
    JimDawson and TomS like this.
  24. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    So I had some time today, and I found I had a bar of 3/4" 12L14 steel. It seemed like a lot for a one time tool, but it kept calling me.

    Ver2.JPG
    Works much better.

    Note to self, if you're going to work with a material you've never machined before, like that Delrin bar stock, get more than you need. "Two is one, one is none".
     
    JimDawson likes this.
  25. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    The last thing I wrote about moving on with this project and cutting the oil grooves was that I had problems with my Sherline/A2ZCNC mill losing motion on the X-axis. My conclusion was that since there's really only a few parts here: the coupler, 1/4" pre-load nut and 5-40 screw, I ordered all of them so I could rebuild the axis as new. I ordered these from a place I've ordered from before and had no problems with. Unfortunately, they royally screwed up the order. I had ordered two of everything: coupler, pre-load nut, 5-40 screws (2 packs of 2). I received four couplers, no pre-load nuts and two of the 5-40 screws. So I emailed them. "Not a problem, we'll get the right parts to you ASAP". The second package arrived and I found two more couplers, no pre-load nuts, and ONE of the 5-40 screws. So now I had six couplers, three 5-40 screws and no pre-load nuts.

    I gave up. I examined the pre-load nut under magnifiers and decided it was OK to work with.

    I had also ordered a 5-40 tap off eBay to clean up the threads on the inside of the lead screw, and cleaned up the 1/4-20 threads on the outside of the lead screw with a threading die. I stuck the end of the leadscrew and the pre-load nut into my ultrasonic cleaner I use for reloading, and cleaned them thoroughly. Ran a toothpick into the 5-40 hole until it came out clean, then cleaned up the 5-40 threads with the new tap. It really only had to work for the first turn, maybe two, although the screws I had in there looked damaged at their ends.

    I rebuilt this on Tuesday afternoon using blue LocTite on the taper of the leadscrew and set it aside for a few hours (which ended up being Wednesday morning). On Wednesday, I started slowly moving the X-axis and all was fine. I had written a simple G-code file to exercise the X-axis, moving it from -4.0 inches to +4.0 inches and back several times. The table stopped moving in less than five minutes.

    I took it apart, of course, and found that the LocTite had sheared. All I did was just reassemble it with blue LocTite on everything. I put a light smear on the taper, a small drop on the 5-40 screw and another drop on the 1/4-20 threaded lead-in threads on the leadscrew. Then I went to LocTite's website to look up just how long it should take to cure, and found a data sheet. The answer was that it really takes 24 hours to get to full strength, so I said I'll leave it until 24 hours is up, Thursday around noon. Again, I started slowly and moved the X-axis back and forth for a while, then ran my short exercise file. This time it worked. Then I ran an extended version, which took about 45 minutes. All in all, I ran it for over an hour with no issues. After an hour of doing things it wasn't capable of doing at all a few days ago, how long does it take to get confidence it can keep moving? Put it this way: if one of those parts had snapped in two, would this even be a question? There would be a broken part and a new, unbroken part, and the only question would be whether it was put back together properly.

    Thursday night and Friday featured Christmas parties, so I didn't get back to it until Saturday. This time, instead of my X-axis torture test, I wrote a little exercise routine for the mill to run all three axes simultaneously for a half hour. Everything held together just fine, so I moved the Grizzly's cross slide back onto the little mill's table, zeroed the coordinates to match the drawing, and then proceeded to cut the slide. It did hiccup, the table started shaking too much. It turned out that when I put the X-axis gib strip back in, I didn't seat it properly. That didn't show up until the X-axis was carrying the full 10 pounds of load. Once I shut everything down and reassembled that last detail properly, everything was fine.

    Groovy-FinalPass.JPG

    This is the last pass on the last groove, and everything worked out fine. The Sherline mill is well behaved, again.

    Now it's on to trying to figure out how all this oiling system stuff works.

    And remaking that ballnut removal tool. It didn't work out, then came apart on the lathe. I'll talk about that some other time.
     
    JimDawson likes this.
  26. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    I came here a few weeks ago to update things, but then got distracted by the holidays, and some of the parts of life that don't involve the shop.

    The first big thing to report is that I got over the operation that scared me the most: drilling a 1/8" hole 4" deep into the cross slide and intersecting another 1/8" hole to bring oil from that hole to the surface oil grooves.

    I had concluded that I'd drill a larger diameter hole and only finish out the last inch or so with the 1/8" bit. There's no reason not to use the recommended tap drill for the larger first hole - it's going to be tapped for the oil fittings anyway, so I settled on using that. Some tables say to use a Q lettered bit, others said R. I ended up drilling the 4" deep hole with R, 1/4 and 1/8, but the 1/4 wasn't really necessary. The cavity from the bigger bits will fill with oil and the only difference is the larger amount of oil in the larger R drill profile. I drilled most of the way using the mill as a drill press, clamping the table to a vertical fixture - pretty much an angle plate.
    LastHalfInch.JPG
    The last inch was drilled by hand since it was too deep to use the mill, and this is it when the drilling was just completed. The 6" long 1/8" bit wanted to flex and move around more than I was comfortable with, so I turned that little piece of aluminum you can see in the hole with a 1/8" hole centered in it, to help me keep the bit from wobbling too much. Once this was drilled, if everything was cool, I should be able to shine a flashlight in the side hole that this long hole is supposed to intersect with and see light.
    DownLongHole.JPG

    With the long hole out of the way, I could work on adding all the fittings, but they were on order. I figured I'd move on to the ballnut removal because that's the main thing stopping me from putting the system back together. I had a bad experience with the ballnut removal tool (which I'll call the BRT). Briefly, I had made a tool out of Delrin and ended up tearing it up trying to shave off some a section of the wall down about .015 thickness. I only had the one piece of Delrin, and since the wall thickness was the problem, I thought I'd make a BRT out of some 12L14 steel I have in my junk box. I ended up tearing that up on the lathe, too.

    While trying to figure out a way to get a BRT that worked, it occurred to me that if I had a piece of the threaded portion of the ballscrew, I wouldn't need a tool. I could just unscrew the nut onto the small piece and then re-screw it back into place on the real ballscrew once it's in the base. All I needed to find was a 5" long piece of RM1605 ballscrew. I asked on Hoss' forum on CNCZone, and was told by one of the guys (not Hoss), "You're way over-thinking this. When you buy a ball nut separate they come with a cardboard tube inserted into the nut to keep the balls in place. " Cardboard tube? Like this?
    BallNut.jpg

    The writer went on to say, "Without having the original cardboard tube, I have simply taken a piece of paper, formed it into a tube, and then wrapped it with masking tape to build the thickness and diameter to where it needs to be and then slid the ball nut off of the screw and onto it". I thought one of those terrible thoughts, "how hard can it be?" In the engineering world, this considered on a level of pure badness about equal to "all you gotta do", which is the most horrifying phrase in all of engineering, and usually precedes the worst times in your career. Indeed, "how hard can it be?" preceded the worst moment in this story.

    Realizing I only need about a .012 to .015 wall, I wrapped three turns of printer paper around the leadscrew, ran masking tape down it, then tried to unscrew the nut. Didn't work. It started binding and when I unscrewed it back off the paper tube, I could see the same problem I had with the first two tools - the balls were trying to create matching screw threads in the paper. So I pulled most of the tape to make the tube thinner and started again. This time it was unscrewing onto the paper without binding. I guess I thought, "holy crap! it's actually going to work." and kept going. A few seconds later, I heard the balls falling into the folded over paper end I was holding in my left hand. At this point, it was over. I unscrewed the leadscrew the rest of the way and found that the last few threads in the nut were exposed and the balls fell out of that. I think I caught all the balls and none escaped, but I won't know until I put it back together. I put tape over both ends of the ball nut and paper cup (folded over paper tube) so that the balls don't escape.
    Ball_Screwed.JPG
    Since I wasn't sure I didn't lose a ball bearing, I ordered some from a seller Hoss recommends on eBay.

    Meanwhile, my oil fittings had come in and it was on to preparing the cross slide for the fittings, which brought its own problem: a bad tap. I had a 1/8-27 NPT tap in a my Horrible Freight tap and die set, and quickly found the cast iron in the table cut away the tap faster than the tap cut threads! After some searching, I found a better tap locally (Ace Hardware) and was able to resume working.

    Here's a view of the cross slide, right side, before tapping for the oil fittings:
    Slide-Right.JPG
    The two holes for fittings are the big hole upper right, and one between the two big cap screws, but below the bottom of the section those screws are on. This next view shows the one in the middle of the front, too.
    Fittings.JPG
    The two facing to the right are straight fittings and the front fitting is right angle. I will use Tee connectors to help distribute the oil around between fittings. The X and Y ball screws have press on, "barb" fittings, and there's another right angle connector on the left side that has the same function of the straight one on middle of the bottom of the right side of the slide. Eventually, there will be six oil tubes on this side and the whole thing will look pretty similar to this:
    OilSystemTubing.jpg
    This portion of the modification is on Hoss's open web site, not on the copyrighted DVD that I bought, so I assume I can link to it. You can see this picture is a screen capture from a YouTube video here.

    The cross slide is now set for the oiling system, so I need to drill and tap the channels for the Z-axis and headstock area. I should be able to take that off the base and finish that once the X and Y axes are moving.

    Which brings me back to the problem of repacking the ball bearings into the ballnut. It's harder than it looks online. That video gives me hope of being able to re-pack the balls without needing a BRT. My problem is that I need the BRT for about 5 minutes. If I rebuild the nut on the screw, like most repacking videos show, I need a fancy BRT to remove it from the ballscrew. But the nut is off the ballscrew already. If I could rebuild the ballnut on something simpler like a piece of doweling or something to hold the balls in place, then put the ballnut onto the screw, I'd be better off.

    And there you go. Just about a full month in the shop. It's a lot but not enough. I want to get this project completed.
     
    brav65 and TomS like this.
  27. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    It has been a good ten days. Lots of progress and the Z-axis is actually running under CNC control - without the headstock and lying on its side.

    This is picture intensive, so bear with.

    I solved the ballnut removal tool problem I talk about in the last post and got the ball nut repacked. I had studied a YouTube video on repacking the ballnuts a few times and tried repacking the balls. It didn't work out. This guy stresses that there are races in the ballscrew that shouldn't get balls put in them, saying if we did, the ballscrew would jam at some point. I couldn't see what he was talking about from the video. About the 10th time I watched trying to see what he was referring to, I seized on a different approach to the job. Feed the balls into the returns and let them push their way around the races themselves, then I don't need to worry about avoiding a spot. The next morning, after suitable amounts of coffee, I put on my Optivisors, turned on a back light flashlight, and mixed the spilled 100 ball bearings in grease. When you watch this video, or many others, you'll see the demo using Vaseline to hold the balls in the race, and it really does work (well, I used Teflon loaded grease - this guy says grease is better but uses Vaseline). Next thing I knew, within an hour's work, I had all the races full of balls.

    Now comes the hard part. When I said repacking "didn't work out", I had tried my previous two attempts at a BRT and the balls got pushed out of place. After that, I decided to go the route of the cardboard tube BRT, but turned one out of wood. I had a "pen blank" (pen blank is eBay talk for scrap wood cutoff) I had bought years ago, 3/4" square and about 6" long. I turned it to slightly over 0.5"; it measures .510 in most places. I drilled a 5/16" hole down its long axis, then worked up the nerve to enlarge that to 13/32 (.406), in case I needed to try to use it on the ballscrew, not just to retain the balls while I screw the ballscrew into the nut. It worked! This BRT (and perhaps a bit more care) didn't displace any of the balls, and I was able to screw the ballnut and mount onto the Y-axis in the mill, gradually (and gently) pushing the BRT out the far end.
    Y-axis-ballscrew-nut-mount.JPG

    Next, I put the cross slide in place with both ballscrews.
    CrossSlide&bothScrews.JPG
    The X-axis screw is just lying on top here. And I needed to cutout that slot around the oil fitting along the way. Now, it's time to check out how far the Y axis moves. It moves 6-1/2"... wait ... Grizzly specifies the original mill at 6-7/8". I lost 3/8"? Back to the DVD videos and finally a quick check with Hoss. The base of the mill needs to be modified to cut away some metal in the front. This is 1-1/4" on a side.
    Pocket1.JPG
    Next step was to cut that out, which I did by drilling a series of quarter inch holes around the lines and then cutting out the webs between them. I used an AC powered drill after my battery powered drill just seemed weak. Amazing difference. I sometimes forget why we used to always use electric tools. My drill has a rating label that says it's 1-1/4 HP, while the battery drill doesn't say anything about power.
    CutoutPiece.JPG
    I cut the webs on the long side with a battery powered reciprocating saw and the short side with a jig saw. I couldn't get the bigger saw in the crosscut space.

    The next step was to take the motor and headstock off. This ends up weighing 50-ish pounds (rough guess). I found a guy on YouTube (N1BPD) who had a video of taking the motor and Z column off and then another that showed him building it back up.
    headstock.JPG

    Which leaves the base and column as the only pieces on the tool cabinet/stand.
    Base&Column.jpg
    On to taking off the column and here's where I hit a dead end. There are four large socket head bolts in the base of the Z column holding that cast iron in place. They take a 10mm Allen key, and as luck would have it, I had two options: a ratchet set from Sears with the novel feature that most of the sockets will pass any length bolt through them, and a nice, tool steel, Allen wrench set's 10mm key. Two of the bolts were moveable and I got them loose, but the other too resisted everything I could do to them. At one point, I actually tried to stand on that 10mm key. Since it was Sunday, I said Uncle and went to think about it and figure out how much I was going to have to pay to break those two bolts loose, while watching the Packers/Cowboys playoff game.

    It ended up being fairly cheap, because after trying to figure out which impact wrench to buy, my wife convinced me to try an adapter to mount the hex key socket on my 1/2" drive breaker bar. Thankfully, that worked and I was able to pull the Z column. This was right after I had watched the video where N1BPD built up his Z-axis and got it running. I decided I had all the parts and had been basically ready to do the modifications on this for quite a while, so why not do it and get the system running? I shot this during installing the Z-axis ballnut getting things lined up.
    Installing.JPG
    Here's a closeup of the motor mount.
    MotorMount.JPG
    If you look carefully at this picture, you'll see something strange. There's a 1/4-20 nut on each of the round, threaded standoffs. When I first built this, the motor quickly jammed and during troubleshooting, I realized the motor wanted to be about an eighth inch farther away than the standoffs allow. A quick check of the relevant drawing said they were supposed to be 2.000 inches long, +/- .005, and they were. Rather than make new standoffs, this was a quick fix. It passes a 10-32 screw to mount the motor and looks neater than a stack of washers.


    You'll notice there's a handwheel on the motor. That's a spare I keep around from the Sherline and use on occasion with it. I used it here to help line up the motors and mounts while getting everything running. It's already off the machine. Everything is a temporary setup, including LinuxCNC, which is using my Sherline configuration file.

    I could have left this for later, after the X/Y table is done, but it's nice to see real progress like this. Because motion!

    I swapped the metal tray on top of the base/cabinet that ships with the G0704 and put in Hoss' big wooden drip tray. I had always envisioned trying to hold the full mill, or parts of it, over the chip tray with a crane and struggling to get it in position. Broken down this far, it's just a 50 pound-ish hunk of cast iron to set down in the tray, finagle into place and bolt down. I think I could use to seal the base so that cooling fluid can't run under it.

    The really amazing part is that I can really see this going together in a few days. Add a solid week for "now where the *&$% did I leave those screws" time-wasting fuster clucks.
     
    brav65, TomS and jbolt like this.
  28. Boswell

    Boswell United States Hobby Machinist since 2010 H-M Supporter-Premium

    Likes Received:
    87
    Trophy Points:
    28
    City:
    Cedar Park
    State:
    Texas

    -Return to Top-

    Wow, you made a lot of progress and had to solve some tough problems. Sort of makes me miss my G0704.
     
  29. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    Thanks. I started working on this project before I retired in 2015 and it just seems like it's taking forever. It has been a constant learning project, which is good because I have little practical machine shop experience, but I really want to be done so that I can use it to make things. It would be nice to just have things go together with no screwing around.

    I started down the path that Hoss calls "phase 1", which is to to put motors on the existing leadscrews. There's much less tearing stuff apart. Once I had all the motor mounts made, I was doing a lot of reading here and other places, and decided to switch to the ballscrew approach, phase 3. That was last March, and threw out almost all of the work I had done to that point. The phase 3 parts are bigger and more complex than the phase 1 parts, so I had to split making them between my Sherline CNC setup and the G0704 itself.
     
  30. MontanaAardvark

    MontanaAardvark United States H-M Supporter - Premium Content H-M Supporter-Premium

    Likes Received:
    58
    Trophy Points:
    28
    Location:
    Just South of the Kennedy Space Center
    City:
    Melbourne
    State:
    Florida

    -Return to Top-

    A friend of mine says that when you do a big project, it's never the big stuff that slows you down. Those are well thought out, well designed, planned and executed. What slows you down are the little things. You know the saying, "the devil is in the details"? So from my optimistic, "The really amazing part is that I can really see this going together in a few days." it has been nothing but little details.

    After getting the Z-axis moving, I went to look at how to get the oil fittings hooked up and found to my horror that I had installed the ballnut pointing the wrong way. There's a hole in the top of the Z-column that allows feeding in the oil tube, and it was on the opposite side from the fitting. That meant zero chance of getting a tube onto the fitting through the hole, or doing it without taking the column apart. After some agonizing over that, I took the motor and mount back off, then pulled the ballscrew and nut, finally putting the oil fitting into the place it goes. That was a step backward. It doesn't fit.
    Z_Fitting2.JPG
    Looks good, right? Until you look in that middle screw hole and see solid metal and not the threaded hole it's supposed to line up with. When I rotate the ballnut to line up the holes, the fitting hits that casting wall. This turns out to be a well-known problem, but not one that has an agreed upon or easy fix that everyone does. There's a seller on eBay who sells the motor mounts (BDTools) that resemble Hoss' plans but aren't exactly the same. He simply screws the oil tube into the tapped hole for it, and then routes the other end to a fitting sticking through the vertical slide. He clamps the tube to the ballnut mount to keep it from pulling out of the hole in the ballnut. Since the tube is 1/4" OD and the hole is M6, which is slightly smaller, it probably would screw in there.

    While I think about how to solve that, the Z-column is on a bench in the garage and the mill is a few feet away, so I went to hook up the oil distribution. Naturally, I found a problem; some parts I hadn't ordered, but one was in town and the other here in 2 days.
    OilDist_Final.JPG
    Now, with the tubing in place and everything else ready, I could put on the mill's table. Which is when I found the next step backwards. The table was rocking left to right and didn't feel right. Turns out it was on top of the tubing from the Y axis (top center of picture).
    Y-axis-fubar.JPG
    I pulled the tube and you can see the issue: the vertical edge on that part of the table goes over the oil fitting. The tubing is almost completely under the slide.
    Overlap.JPG
    I tried bending the tube over so that it comes out minimizing contact with the slide. Polyurethane tubing takes a permanent set very easily with a little heat applied. The bent tube still hit the slide. Then I switched to a smaller diameter tube. Heating the tube and shoving it onto some needle nose pliers allowed it to fit on that fitting, which is really a size too big for the tube. That worked, but it still rubbed on the slide, so I'm trying a more desperate solution. I enlarged the tube to press on, then bent it at a hard right angle so that it goes straight down (as seen in that picture) toward the middle of the cross slide. Right now, that's sitting glued down with RTV and under 20 pounds of weights until the RTV sets.

    If that doesn't work, I see only one thing to do here. (Well, two things, but one is to machine away the interfering part of the slide and I'd need a mill at least as big as the G0704 to do that. ) I think if I take the Y axis ballnut mount off, that's the shiny aluminum with machining marks you can see there, and extend that cutout around the oil fitting all the way to the wall of the cutout it's sitting in, that would allow me to put a right angle fitting in its place, and bring the tube out of the cutout straight toward the middle of the cross slide and never touch the slide.
    G0704 C7 Y axis ballnut mount angled no chamfers-modified.jpg
    It would be removing that material marked in red. It makes the cutout extend all the way front to back, instead of just down to the floor of the part that the ballnut rests on.

    The downside here is I have to completely take the base apart, again. Remove the table, take off the steeper and all the drive hardware on the Y axis, take off the cross slide, put the ballnut and mount back on the ballnut removal tool by removing the ballscrew, take off the ballnut mount, and machine that slot on my Sherline system, then rebuild everything. Lots of these screws and fasteners are LocTited in place. It would be a couple of days worth of work. Which probably means another week. On the other hand, this hose is completely invisible during operation and if it wears through, the first sign would probably be oil leaking and oozing out from under the base while making something. If gluing the tubing down doesn't fix this, it seems like it has to be done.
     

Share This Page