Big CNCs

Jump to navigationJump to search


We've got 2 big cnc's underway, the Guy^2 and Stephanie. Some files on how to choose motors and a load-calculator:




we also got a mini cnc


Guy & co brought over an almost-ready CNC machine he's been building with two friends, on-and-off for the last two years, as a loan for the purpose of getting fresh minds and hands working on it and completing it. If we get this beast up and running, we'll have a major addition to our arsenal!

The CNC is professionally built from large aluminum profiles custom-cut to size, as well as some CNC machined parts. It has brushless servos with encoders on each axis with matching drives (400W for X and Y, 200W with brake for Z). The spindle is a 2kW water-cooled beast with a standard CNC collet. Motion is on accurate linear slides and driven by precision ballscrews.

Note: a LOT of time and money were invested in this machine and it's quite easy to do expensive and/or dangerous mistakes. Therefore, if you want to help, please consult first with the people already working on it and they'll get you up to speed - what was done, what has yet to be done and most importantly, what NOT to do... At the moment (4-3-13) that means talking with Guy O. and Kfir.


  • wire the limit switches to their respective control board (X, Y and Z - 6 limit switches). they should be wired to pins 4 and 5 ("P-OT", "N-OT") on the big connectors of the respective drives. enable those inputs through Smartjog, test if they work (also via Smartjog, the "digital monitor" section).
  • wire the limit switches to the indexer board and connect to mach-3. We will use a single parallel port pin due to pin shortage (port has 5 input pins, we need 2 for each axis, i.e. total of 6). in order not to short the switches to one another, so the servo drives still have individual indications, we'll use diodes when shorting them to the one parallel port pin.
  • the Z axis motor has a brake, that works on 24V. it is currently hard-wired to the 24V PSU. ideally, it should be connected through a relay controlled by the Z-axis servo drive (the drive has a specific output to control a brake, but it's low power - i.e. an external relay is required).
  • test water connections and pump and gantry for leakage. seal as needed.
  • continue tuning the servos (when connected to the CNC) for better performance. Cation - incorrect tuning can lead to dramatic, dangerous and out of control machine movement.
  • find a better location for the electronics box (better: re-do the electronics box - but that's better left for upgrade work after things are already working)
  • get a proper computer for the CNC, instead of the inconvenient laptop. It must have a physical parallel port and serial port (either natively or as an expansion PCI card). It doesn't need to be very fast but it should be decent. put it on the bottom shelf of the table and wire it to the CNC. install windows (XP or 7), Smartjog, Mach3. dual-boot with linux is also possible (for LinuxCNC instead of Mach3).
  • configure Mach3: set axes properties, acceleration, etc.

  • temperature sensor on the outgoing water from the spindle.

some pictures: guy's cnc guy's cnc guy's cnc and newly milled piano attaching plate to table


  • Z-axis was fixed
    • There was a wiring problem of the Z-axis that might have done damage to the servo driver.
    • servo driver was replaced (new, same board)


as of july 2013, STEPHANIE was Disassembled, with great sorrow.
Short Circuit - No Disassemble!

We have a 180cm x ~100cm alum. plate 1cm thick holding some green x-rails bolted to the table on top of some unit profiles. A rack runs the length of the rails to connected to pinions on the y stage motors. The pinion we do have is 30mm diameter (to middle of teeth) and there are 32 teeth so the metric module is 1, the gear being referred to as 'module1-32'. Check calculation. For a smaller gear that fits the same rail you get a module1-20 for instance, for 20 teeth at the same 'pitch' or teeth per (azimuthal) inch.

Kfir has jumped on this one. We reconstituted a table for the cnc and I sliced off some length from the rails. Now we need to attach rails to table hopefully with modular 8020 or unit profiles so height can be adjusted (to change lever arm for large work pieces (tall) or heavy milling (short)).

Z stage

Z stage Z holder back Z holder front Z holder assembly tool holder brackets

pinion on motorpinion on motormotor bracket x axis 'works' x-axis left x-axis right motors arrive - nema34 8Nm attaching plate to tablein the beginning xaxis Coolant.jpg older pics here

Base - aluminum plate about 1mX2mX1cm

moving gantry with X,Y based on rack+pinion, and Z stage prebuilt

A dremel or other tool gets put on the z stage


  • make/obtain 80V 20A power supply [1]
what details source price
wound toroid 80V 1kVA salame 33 500nis
wound toroid 75V 800VA antek $84+$50ship
wound toroid 75V 1.5kVA antek $155+$50ship
moti's ups 72V output ok, need more caps moti 0
isolation xformer 110VAC 2-3kW ? ?

  • Get build motor drivers - need 6A 80V x3motors.
  • Get longer linear rails for y (1meter X 2 or 3 units)
  • bore out gears to fit 14mm motor shaft
  • Attach Z stage to Y stage
  • Get dremel, attach to Z stage
  • add endstops/limit switches
  • add setscrews to gears
  • Get 12mm SST threaded rod ; need 4x25cm=100cm
  • make table/legs to hold base
  • cut 4 or more aluminum profiles to same length, possibly at the aluminum supply place near us
  • Attach rails to base
  • Get some Al plate 0.5mm thickness
  • Get more Al plate 0.5mm thickness:
    • 40.5cm x 15cm x 0.5cm for zaxis
    • 10x10.5x >= 1cm for block holding z axis to y axis
    • 10cm x 10cm x 0.5cm L profile for x motor holder - try arad
    • 10mm x 8cm x 1.3m for future use
  • Attach x-racks to rails
  • construct x-motor holders
  • Attach X motors to Y stage
  • Get rack for y
  • Attach y-rack to y-stage
  • Print adaptor ring between stepper and pinion gear (from 6mm to 10mm diameter iirc) NOT NEC.
  • Print motor holder NOT NEC.
  • Attach Y motor to Z stage


what details how many
Stepmotor driver 6A ~80V (from 6.5mH and Vm=32*sqrt(L[mH])) 3
pinion (spur) gear {order in progress} module 1 ~20 and 32 teeth, large shaft (>20mm), bore 14 or as close as poss. w setscrews 6
nut for acme screw 2 starts, 11 tpi 1
nut for acme screw The big one (tpi? diameter?) 1
m4 Allen no head 10mm 10
m6 bolts >=120mm length 7
L profile 10mm 100mm sides 1m
Stepmotor 6A NEMA34 1232OZ-IN MOTORS ORDERED 3
Al block 40mm X 100mm X 160mm 1
Al block 60mm X 100mm X 160mm 1
drill bit 14mm for sst, with shank <=13mm 1


link details price
[2] 45V 6.3A $79x3+31 shipping
[3] 70V 6A $54x3+45 shipping
[4] 70V 6A $75X3+0 shipping
from guy 7A 60V $50x3 no shipping to IL
from kfir 7.8A 80V $189 +215 shipping

DIY driver

we got 10 HIP4081 samples : File:Hip4081dat.pdf,File:Hip4081an.pdf , these are 80V hbridge nmos drivers.

The first step is figuring out all the external component values. I arrived at this without delving into the formulas, just using the example from the dev board. Hip4081values.jpg The values for Cx, Cy are not listed anywhere I can find so I guessed 0.22uF.

We have a bunch of onsemi MLP2N06 transistors which are 62V 2A devices, a little low for current app but enough for a start. Also the irf540

We also have vishay Si9978 chips, 40V max Hbridge nmos drivers. We'd like to put these onto a board layout with high power mosfets using eagle and print a few boards for our own use and possibly for others.

good general hbridge info here Guy and kfir managed to hand solder the smd vishay chip to this breakout board Yair brought, with any luck we can get an H bridge running this wk.

To translate step/dir into individual coil activation a simple logic ckt can be used as in File:Stepdir.pdf and here as odt

dc power supply

would like 60V and something like 20A requiring an xformer of N~3 and big caps of ~100mF.

circuitsim online Dcpower.jpg Dcpowersim.jpg