pic: CIM Printer Servo

This is a CIM motor setup as a DC servo for a large-scale 3d printer I’m working on at the University of Arizona College of Architecture. The CIM, US Digital encoder and GeckoDrive servo driver just arrived in the mail a few days ago and I finally had a chance to try them out. Needless to say, the precision is fantastic - I never knew a CIM could move so smoothly (good robo-pickup line).

The printer itself is designed to print a concretious material (currently looking into fiberglass reinforced concrete). The print bed is roughly a square meter at the base with about .5 to .75 meters of vertical travel. We purchased the GeckoDrive and CIMs for the x- and y-axis after we hit the physical limitations of the 250oz-in steppers origonally used.

(I do not endorce duct tape encoder mounting method.)

Very interesting use of a CIM. Could you provide more info on your hacked servo setup? I’m assuming you used a G320X. What encoder did you decide to use? What DC supply voltage are you using? Your creativity has the cogs in my brain spinning! Thanks!

It is indeed a 320x. We purchased an 8mm encoder from US Digital (http://www.usdigital.com/products/encoders/incremental/rotary/kit/e7p) which fits perfectly on the shaft. We’re using an old 24v power supply that was laying around which can source at least as much current as max the 320x can supply to the motor, which is 20 amps. I probably should have spent more time looking for a more suitable motor (higher voltage, the 320x can take up to 80v but the CIM can draw upwards of 100 amps, so it’s operating at a relatively low torque), but $25 is a heck of a deal! I’ll be mounting it today and hopefully will have at least one axis working. The motor shouldn’t hit the driver’s current limit if the gantry is accelerated slowly enough, but the hacked firmware we’re using (http://reprap.org/wiki/Sjfw) supports adjustable accelerations on each axis.

The current motors only have a holding torque of 250 oz-in, which decreases sharply with speed, and, as I’ve had the opportunity to learn, steppers need to be accelerated very slowly to reach the speeds we need (at least on such a large scale). Currently they’re being accelerated at 30mm/s/s up to about 10cm/s, but we’d like to at least double the maximum speed and reduce the amount of accelerating travel.