By popular demand, a schematic which roughly shows the electrical connections of our cnc-mill setup. Most of it fits inside one box (also shown here). The m5i20 I/O connectors are on the left, followed by the optoisolator cards. E-stop chain in the middle, servo-amplifiers and motors to the right. Jog wheel at the bottom. The small VFD board I made later is not shown.

I’m fitting some more equipment around a Nikon TE-2000 microscope, and the stock objective turret is in the way. Machined this objective holder in steel yesterday.

Here’s a setup wit three vises for machining model yacht tiller arms. The parts are rotated 90 degrees between the first stage (leftmost) and the second stage (middle), and then again 90 degrees for the final stage (right). There’s some rigid tapping at around 8:20.

The motor is a standard induction motor from ABB rated at 1.5 kW and around 3000 rpm (at 50 Hz AC). It has a lot of model identification numbers: “1.5 kW M2VA 80 C-2 3G Va 08 1003-CSB “. There are more details on this line of motors on ABB’s site, but this kind of motor should be available from almost any manufacturer of industrial induction motors. Presumably torque drops off after the rated maximum rpm of 3000, but with small diameter tools we have been running the VFD up to 90 Hz or around 5400 rpm. When taking heavy cuts the VFD tries its best to keep the rpm up, but we do observe a 100-200 rpm drop when a 40 mm face-mill digs in. It might be possible to wire the encoder counts to the VFD and get a truly closed loop system but I doubt it’s worth it.

The motor is connected to an Omron Varispeed V7 VFD with a maximum motor capacity of 1.5 kW. I can’t find a good link on the international Omron site, so here’s one in finnish instead (datasheet here). This is a sensorless vector-drive VFD, which is very important - with the previous simple V/f open-loop VFD we could only do machining close to maximum RPM and certainly would not have tried rigid-tapping. The electrical connection is simple with the VFD connecting to 1-phase AC mains and then the three phases from the VFD connecting to the motor.

The VFD is controlled by EMC2 using three general purpose IO pins on the m5i20. One pin sets the rpm (VFD reference frequency) using a pulse-train generated by the stepgen HAL component (step_type=2 ctrl_type=v). The two other IO lines set the VFD to either forward or reverse.

On the hardware side of things there is a 1:1 belt drive to a littlemachineshop MT3 spindle (more here).

Below a close-up of the US Digital 500 ppr encoder mounted on top of the motor. There’s a cooling fan driven by the main motor axle on top of the motor and we tapped the axle with a M6 thread, inserted an M6 set-screw, and coupled the set-screw to the encoder using plastic tubing. The encoder sits on a alu-bracket which is bolted to the fan grille. Z-axis servo in the background.

We’ve mounted a 500 cpr encoder on the spindle-motor which means it’s possible to do rigid tapping. Above some spot-drilling, then a 2.5 mm drill, and then an M3 tap at 500 RPM and 0.5 mm Z-feed per revolution. Below the same thing but with a 5 mm drill and an M6 tap (1 mm Z-feed per rev).

Cool stuff!

Last weekend we made some rings for mounting a finderscope on the main telescope. Today I got the holes drilled and tapped so the rings are ready to use. The gallery below shows the adaptive pocketing paths that were used to cut both the part and the jig that allowed drilling the holes at +/- 120 degrees.

There is a video of the roughing operation at 2000 mm/min that will be online soon.

For a mostly home-built cnc minimill of this size, this is pretty ‘hard core’ stuff…

After the test run on Saturday Jari made a complete bulb in steel on Sunday. The first half can be milled with the stock clamped to the vises, but for the second half we need this jig. It’s in aluminium and was fairly simple to make - which also means making a bulb mould in aluminium should be easy. If someone is interested in a bulb mould, do drop me an email.

Milling the second half proceeds exactly like the first half. Here the rough-program is run leaving about 1 mm minimum of material for the finish pass. We now adjusted the program for a bit faster feedrate and much faster plunge-rates as it is clear the program is error free and all plunges are outside the stock.

Surface finish is slightly better than on the trial bulb. The design weight was 2410 g and this one came out at 2416 g - pretty good. With a 100-150 g fin trimming the total weight close to 2500 g shouldn’t be a problem.