For a while ago, I wanted to create a big cnc router. So I needed some tools to build this. Therefore I bought the HBM250, which is a mill-turntable combination. Because that's not very handy, I bought a additional XY table, for mounting the BF16. Later I also bought some stepper motors and thought, why creating a router? I can also rebuild my BF16! So from this year (2009), the build progress could by followed on this page. This mill is sold onder different names: Opti BF 16 VArio, Quantum BF 16 Vario, HBM BF16 and in the states: XJ-9510. (There are some minor changes in spec. like max. rpm 2250 to 3000.) EdgE's Home

Properties BF16

2008
Some properties:
  • Travel X: 240 [mm]
  • Pitch X: 3 [mm/rev]
  • Travel Y: 170 [mm]
  • Pitch Y: 3 [mm/rev]
  • Travel Z: 210 [mm]
  • Pitch Z: 4 [mm/rev]
  • Spindle travel: 50 [mm]
  • Spindle speed: 50 - 2250 [rpm]
  • Spindle size: MC 2
  • Spindle power: 500 [Watt] (0.75 [Pk])
  • Spindle weight: 60 [kg]
  • Table size: 460x180 [mm]
  • Table T-slots: 3x 10 [mm]
  • Table weight: 80? [kg]
  • Stepper size: NEMA84
  • Stepper holding torque: 12 [Nm]
  • Stepper weight: 5.3 [kg]
  • Stepper controler: MSD980
  • Stepper power supply: PSU 70 [Volt]; 5 [A]

    • Resonances

    		june 2009
    The BF16 is a very light milling machine. So vibrations are very notacable and could possibly optimized. Sources for resonaces:
    • Play of leadscrew (backlash) & play of guidance:
      • X axis
      • Y axis
      • Z axis
      • Spindle axis
    • Stiffness total construction:
      I will do some measurements and calculations later.
    • Stiffness tool.

    Construction Stiffness

    		July 2009
    I did some analyses on the vibration of the BF16. I'm not sure what kind of cast material is used, so a guess: Iron Cast G40 - Emodules of 140 [GPa]. Here a list of the first 6 modes:
    Mode 1 - 31 Hz
    		 Mode 2 - 45 Hz
    		 Mode 3 - 170 Hz
    Mode 4 - 195 Hz
    		 Mode 5 - 390 Hz
    		 Mode 6 - 440 Hz
    It's clear the column is the weakest point of the construction. Also the stiffness is not very high:
    K_x = 3.2 [N/mm]K_y = 2.9 [N/mm]K_z = 6.8 [N/mm]
    Time for some re-enforcement!
    K_X = 9.4 [N/mm]:
    		 K_Y = 43.5 [N/mm]:
    		 K_Z = 47.6 [N/mm]:
    		 Construction:
    Mode 1 - 83 Hz
    		 Mode 2 - 105 Hz
    		 Mode 3 - 230 Hz
    Mode 4 - 370 Hz
    		 Mode 5 - 510 Hz
    		 Mode 6 - 520 Hz
    Conclusion: by adding an additional support, stiffness could be optimezed a lot! An other nice option, bolt the column against the wall.

    CNC Kit

    		April 2009
    Because I earlier planned to create a (big) router (2x1.5 [m]), which should have high speeds, I needed to have big steppers. I was thinking of spindles, with pitches of 20 [mm/rev] or more. So I decided to aim for 6 to 8 [Nm] steppers. I also thought of building the controler my self, so did some calculation for R/L driver approach, but didn't continue. The electronics are to difficult for me as a mechanical engineer. After searching, the best price/performance, I decided to buy my stuff at MotionControlProducts, 'CNC Drive Kit 4'. In that periode, they where out of stock for the 8 [Nm], so why not order 12 [Nm]? :)

    After some testing, all the wire connections will be bi-polair parallel connected. For 'high' speeds, the driver will only ask for 1.5 [A]. I don't know yet, what current will be used, while milling. If this will be below 2.5 [A], 2 drivers could be connected to 1 PSU.

    Speeds Y-axis (nut 2003) (Y-axis)

    		june 2009
    Finaly I started to rebuild my machine and mounted a 12 [Nm] motor on the Y axis. I'm very proude :)
    For testing the controler & LPT, I programmed simple Visual Basic program. At maximum speed (only 64 [Hz]), the Y moved forwards & backwards very slow, but it works!!
    Now I know, I don't blow up the LPT and everything is working, I have to try different software, which could speed up everything. So I installed Mach 3. The software installation was okay, but to configure the stepper was very difficult. Maybe I didn't read well enough, but in the end I had to change the LPT settings in the Bios. With LPT ECP type, Mach 3 was able to communicate. (VB had still no problems.) I'll post some pictures of settings later. Now ofcourse the fun part starts, how fast could the machine run... These are the current settings for Mach3:
  • Wire connection: Bipolar Serieel/Parallel
  • Step size: 5 or 10 [-]
  • Pitch: 3 [mm/rev]
  • Kernel speed: 25/35/45/75 [kHz]
  • Steps/rev: 200*5/3 = 333.333 (or 666.666 for stepsize 10.)
  • Max speed: 1-3 [m/min]
  • Acceleration: 50-150 [mm/sec^2]
  • Step time: 2 [nsec]
  • Dir time: 2 [nsec]

    With above settings, I also measured the current between PSU and Driver (Parallel wired):
  • Idle: 0.2 [A]
  • Constant 'high' speed: 0.7 [A]
  • Max accelerating: ~1.5 [A]
  • Temperature of motor: ~45 [dC]

    After some testing, see graph, these are the best settings:
  • Wire connection: parallel
  • 35 [kHz]; stepsize 10
  • Max speed: 3 [m/min] (didn't go higher because of low acceleration)
  • Nice speed/acceleration ratio: 2.5 [m/min] & 250 [mm/sec^2] (833 [rpm])
  • Without load (disconnecting coupling): at least 2000 [rpm] (@65[kHz] & 500[mm/sec^2])

    It's a pitty, that Mach3 has a constant acceleration factor. If not, it should be possible to get much more speed out of the systeem, with higher accelerations. But in the end, I've found nice speeds/acceleration ratio: 2.5 [m/min] & 250 [mm/sec^2] will work.
    		•

    Speed Z-axis (ballscrew 1605)

    		October 2009
    Finaly some progress. I did some tests with the spindle (horizontal) and after that, I mounted the Z-axis with ballscrew spindle & 12 Nm motor. This is running very smooth. Also the speeds & accelerations are lot better. Only one strange thing happens. There is a strange peak at 2.5 [m/min] & 2.5 [m/sec^2].

    Power Curve

    		October 2009
    Because of the strange peak, this requires some more investigation. I could only find one graph of my motor, but With 2 formulas it's simple to create your own power-speed & torque-speed curve of your own stepper. You only need to know the rotor inertia and software with a fixed acceleration. Without any mass attached, the acceleration speeds before ~220 [RPM] are very high, so it's not realy possible to measure anything. From 220 to 1000 [rpm], the graph is also very strange. This is not what you expect of a stepper motor! In my opinion, the accelerations are far to high, or are in conflict with the maximum frequency output of Mach3. This is simple to adjust. By increasing rotating mass, the acceleration should be lower, so there are no software acceleration limits. This is also shown in the graph. Now the motor asks about 140 [Watt] and the strange acceleration points are gone :) Both graphs are not exactly the same but gives at least some insight.