• X Axis ballscrew conversion and mounts

    David Pye • 04/23/2017 at 18:52 • 0 comments

    'm currently working on plans for CNC conversion of the X-Axis and installation of a ballscrew and ballnut.

    Like the Y Axis, I plan to use a 20mm 5mm pitch ballscrew.

    At the moment, I don't have the ability to machine things very accurately (existing X axis backlash is horrendous!), I am using flanged ballscrew supports (FK15 and FF15) and paying the supplier of the ballscrews to turn them down and thread them to fit the flanges.

    The flanges for mounting the ballscrews look like this:

    The FK flange has a securing nut to tightly grip the ballscrew so it can rotate, but not move backwards or forwards. The FF flange simply supports the ballscrew but doesn't stop it moving backwards or forwards. The reason it isn't tightly secured at both ends is to account for change in its' length due to thermal expansion.


    For the X axis, as you look at the machine, I am going to fit the FF15 leadscrew support on the LEFT hand side of the X axis table, and the FK15 and stepper motor mounting onto the RIGHT side.

    X axis ballscrew mount - left hand side


    For the FF15 bearing mount, I have designed the following plate (which I am machining out of 3" x 1/2" thick Alu bar):

    This has a centre hole and four M4 drill holes (which I will tap out to M5) to take the FF15 flange. The two slots are for mounting to the X axis table using the existing mounting bolts and will allow some adjustment in the vertical position of the plate.

    X axis ballscrew mount - right hand side

    This mounting is a bit more complicated, and consists of FOUR pieces of milled aluminium bolted together.

    Plate 1:

    Firstly, a plate similar to the FF one, which bolts to the mill table using the existing bolts: This one is slightly thicker, and is made of 3" x 3/4" thick bar.


    Main differences here are:

    Thicker plate (15mm) as the FK15 bearing mount is thicker and it needs to be contained within the plate ideally.

    Four additional holes (countersunk at the rear of the plate) to take M6 allen head bolts to bolt the spacer plates on, which will then themselves secure to the motor mount)

    Plates 2 and 3 - spacer plates

    The spacer plates are 60mm 'high', 10mm thick and the appropriate length to accommodate the shaft of the ballscrew, the flexible coupling, and the stepper shaft.

    They are tapped to take M6 bolts to attach them to the FK15 mount plate shown above.

    CAD files and exact length specifications pending!

    Plate 4 - Stepper mounting plate

    This plate bolts to the two spacer plates and completes the FK15 and stepper mount assembly. The stepper motors I am using are NEMA24, so the hole spacing etc is made to suit it.

    This plate is also machined from 3" x 1/2"thick Alu barstock. It has a mounting hole for the stepper motor shaft to protrude through, a bored out 'lip' that the stepper motor centering 'lip' sits into, as well as four tapped holes for the M5 mounting bolts to hold the stepper in place:

    Once I've machined and assembled these, I will post photos which will make things a bit clearer!

  • CNC control electronics build

    David Pye • 04/11/2017 at 22:07 • 0 comments

    I've had a few days at home on and off to work on assembling the main stepper motor driver hardware.

    The basic architecture is based around a 36VDC PSU, 3 M542T stepper motor controllers, and a opto-isolated parallel port interface board.

    I found these all fitted very nicely into a small PC case I purchased from Amazon, as seen below:

    (wiring still needs to be neatened up with some wire-ties, as well as making a small panel to mount the 4 pin connectors to plug the steppers in). For connectors, I am going to use 4 pin 'aviation' connectors, which are screw-in, thus reducing the risk of motor disconnections which can damage the drivers.

    I have also added a small DC-DC converter to produce a 12V supply from the 36VDC PSU. This means I can drive the original case fan to improve airflow over the drivers.

    The wiring diagram showing how to wire the parallel port interface card is here:

    (Note the stupid design of having a USB 'A' SOCKET for 5v power *IN* from the PC. They've supplied a USB 'A' to USB 'A' lead for this purpose. <shrug>)

    Also, the 12-24VDC supply shown on the right is optional and only needed if you are using a 0-10V output signal for the inverter. I'm not.

  • Y Axis Ballscrew Mountings Part 1

    David Pye • 03/31/2017 at 22:22 • 0 comments

    For the Y axis mounting for the ballscrew, I decided to use a FK15 standard flanged ballscrew mount.

    I chose to drill 6 M5 holes (to tap to M6) in the front of the mill casting to support an aluminium plate with an FK15 cutout (to make later!)

    Because our Makerspace had a laser cutter, I created a DXF file to use as a drill template, and laser cut it out of 3mm ply wood. You could equally print it out and stick it on, or mark the holes by hand (carefully)

    The 6 outer holes need to be drilled to M5, and then carefully tapped out to M6. Because the base of the mill is cast iron, no lubrication is required to drill or tap, as the carbon in the cast iron provides this. I used the M7 bolts that had held the original Y axis leadscrew mounts to hold this plate on to ensure the holes were in the right place.

    Here's a photo with my laser cut ply drill guide bolted to the front of the mill prior to drilling of the holes.

    The DXF file (and the freecad file used to generate it) is in the Projects file section. Note: The file doesn't include the centre holes for the FK15 ballscrew mount (the large centre, and the four surrounding ones) so just has the two holes for the existing bolts, and the 6 edge ones for the new mounting bolts.

  • Software/Electronics/Electrical Design

    David Pye • 03/26/2017 at 01:43 • 0 comments

    Components Needed

    • PC to interpret G-Code and provide control signals to stepper motor drivers
    • Interface board from PC to stepper motor drivers
    • Stepper motor drivers
    • Stepper motors
    • PSU for stepper motors

    I will explain a little about the options I considered for each, and what design decision I reached for each one.

    1. PC - >Like most hackspaces, Makerspace is well-supplied in older PCs, and a suitable Dell PC with a 'real' parallel port was easily found, This is important because a traditional (ie not USB->parallel port emulator) parallel port allows, with the use of real-time Linux kernel patches, precision control of stepper motor timings and pulse frequencies. The PC was installed with Linux-CNC, which is a Ubuntu distribution with preinstalled E3 CNC software and real-time kernel patches to ensure accurate timing and control. E3 is open source CNC software which 'reads' G-Code, and generates, via the parallel port, the signals needed to move the motors. Project website: Linux-CNC Project
    2. Interface board. You don't have to have one of these, but it is generally considered good form to provide opto-isolation between the parallel port pins and the stepper drivers/other electronics. I sourced a '5 axis MACH3 CNC Breakout Board for Stepper Drivers" ' from eBay - it is a fairly generic Chinese clone board, and provides the following features:
        1. Opto-isolated interfacing with 5 stepper motor controllers (X,Y,Z,A,B)
        2. A relay for controlling an external device (eg coolant pump)
        3. A 0-10V signal for feeding a 3-phase motor for spindle speed control (adjustable)
        4. Interfacing for limit switches/encoders etc
      1. Stepper motor drivers - your choice here depends on your choice of steppers. I tried some Linistepper drivers I had to hand but even with adequate heatsinking they weren't able to handle to power needed to drive my motors. Therefore, I bought a set of 3 M542 stepper drivers. These are nice and modular, and already have appropriate heatsinking. They have screw terminals for both motor coils, DC+/GND in, and connections for PULSE, DIR and ENABLE signals from the interface board above. I plan to add a fourth, for a dividing head or similar later on, to enable cutting of gears and other things requiring rotating the workpiece. The M542 drivers are rated for up to 50VDC supply voltage, current up to 4.5A, pulse freq to 300KHz, optoisolated imputs, and able to do a microstepping.
      2. Stepper motors - I ended up with 3 NEMA-24 8-wire motors, rated at 4NM of torque (in bipolar mode), and requiring 3A current)
      3. Power supply - again, after more eBay shopping, I selected a brandless power supply, rated at 36V (600W), and easily capable of driving all the motors/ancillaries I needed.

    3. Basic concept/the starting point

      David Pye • 03/26/2017 at 01:18 • 0 comments

      Makerspace has an RF30-clone mill (these are originally made by a Taiwanese manufacturer called Rong Fu), and rebadged and resold by a large number of third parties. This was very kindly donated by one of our members, Ben Shaw.

      Ours is name-plated as the "MSC Industrial Supply Company" but there are a number of other resellers, including Axminster, in the UK.

      Because of this, there are often incompatibilities between the machines, as each manufacturer has made customisations to their models.

      It is a ROUND COLUMN mill - this means that you can adjust the height of the mill by undoing two bolts and turning the height adjust handle. But because it is a round column design, the quill will lose its' alignment in X and Y, as well as changing height. This is a bit of a disadvantage, but is not trivially easy to rectify.

      Drive system
      It is powered by the original single phase 240VAC 1.5HP motor, and has a belt and pulley arrangement to allow a variety of drive speeds.

      The machine axes are as follows:

      Y axis. Located most inferiorly on the machine. The lead-screw is left hand thread, and drives a brass casting with the lead-thread cut in it. This moves the Y axis along the machine's slides (front to back)

      X axis - acme lead screw, brass lead-nut is bolted to the centre of the Y-axis table, the lead screw is anchored at both ends, and by turning the screw the table moves (left to right)

      Z axis - coarse adjust, and a lockable fine-adjust with a worm gear driven by a Z axis handle.

      Current issues

      1. Most of our members probably think in metric - this machine is imperial.
      2. The X and Y axis brass lead-nuts are worn out, leading to a lot of backlash (this means that when you change direction, and try to turn the table the other way, there is a lot of 'play' in the drive system, making it very hard/impossible to work out how far you have moved). Therefore, accurate drilling/positioning is difficult.
      3. It's a round column mill with the issues described above

      We can address points 1 and 2 by a CNC conversion , involving replacing the leadscrews with a nice modern ballscrew and ball-nut, and stepper motor drive.

      Point 3 is something we will just have to live with!