Circuit Mill Development Log: Z Slip Fixed - Promising Results

Shane Oberloier's picture

The demands of graduate school were taking their toll last month and I wasn't able to give OCI the love it deserves. But I'm back to work now! The Z axis was not getting enough current from the stepper driver on the ramps board. This is because the entire spindle assembly weighs 1kg - and is straining out the stepper motors. When moving in small increments, the motors will skip steps and move inconsistently (see data from Circuit Mill Development Log: Z Slip). Often times the spindle would slip downward while cutting, causing some catastrophic results.

No Z movement (left) vs Level compensation (right)
No Z movement (left) vs Level compensation (right)

At first I'd hoped that it might just be slack in the belts - but that didn't have much of an effect. So the next thing to test was increasing the current to the motors. I chose a TB6600 based driver as the new driver. I could have connected it to the RAMPs board with female header cables - but that wasn't a preferable solution. Instead I utilized my handy Stepper Motor Breakout design, so I could cleanly wire the step, dir, and enable signals. The video below explains the basics of the board.

My connections to the stepper driver can be seen below. It is configured for 32x microstepping, and 1A of current.

Stepper Driver Wiring
Microstep driver all wired up!

After a bit of fiddling and recalibration (see OSE D3D Circuit Mill: Steps per mm Calibration) I was ready to give it a test. Unfortunately the auto leveling procedure is not working properly in OCI Copper Carve, it needs to be redone using a timer based function (that's what I'm fiddling with next). So instead I am leveling the design by hand with my g-code leveling utility. Doing 16 point measurements I was able to successfully mill the Stepper Motor Breakout design and almost successfully milled the RF Transmitter circuit. I attribute the errors to having my backlash compensation disabled. I need to remeasure all backlash values before enabling the function. As a complete side note; I covered the carving surface in a light layer of baby oil and it does a great job of keeping the tool cool (read as - not breaking) and keeping the dust down. So without further ado - here are the current results!


Test geometric patterns and test designs milled with reasonable success