I’m setting up a sort of Frankenstein-CNC where I am using motors and an old motor shield with Pololu motor drivers from my Shapeoko 1, and new body parts from Xcarve, and even a piece of Maker Slide from it’s Kickstarter! But, overall it’s quite sturdy and seems ok.
I flashed the Arduino UNO with the newest GRBL, and I seem to have lost the settings I was using. They were really outdated anyhow since this is a Shapeoko 1 era machine. The new GRBL has a ton more options.
I’m trying to calibrate it with the Universal Gcode Sender Java and everything seems fine. I’m setting up my steps per MM, and I seem to have that dialed in ok. I issue a X10 or Y10 command and measure it with some calipers, and I’m close enough for right now. All Axis test out just fine when issuing manual commands.
So, I grab a simple piece of GCODE and run it, and there’s movement, but just barely. It seems to me like the whole project is scaled. It’s very strange. I tried a couple of samples and am getting next to nothing.
Any ideas or pointers for getting this going, or what I’m doing wrong.
Thanks a million for any ideas.
Upload the G-code file here.
G20? G21 in the header metric sae
Thanks a million for the help you guys.
Here’s a the simple Gcode I was trying to use…
I went back to Shapeoko build instructions since that’s what my electronic were tuned to.
Huh, G20 at the start, which means it’s calling out inches as the format, eh?
Would UGS do on the fly conversion to MM? That sounds like a dumb question.
Would my shield only be configured for mm?
(I have a lot to relearn obviously)
Thanks again for the help!
Hi Robert, this was just a precanned piece of Gcode to start crawling.
I was using an old version of UGS…so that has helped. I got something this time. My Y axis is scalled 50% still for some reason, but it’s working. I’m going to try and find settings for the Pololu(i). Maybe one is using a microstepping that the other isn’t. That might explain it.
Thanks Phil. That makes sense. I definitely know the old UGS was an issue because I got an OK cut this morning. So that’s good. Now I just need to figure out one axis is scaled totally wrong. But I’m gettting there.
Are you still using MXL belting?
For the Z-axis, do you have an M8 threaded rod, rather than the Acme screw which is more prevalent now?
For a general discussion of Grbl configuration please see: https://www.shapeoko.com/wiki/index.php/Grbl_Configuration
Hi Will, I’m using the Acme screw and MXL belting.
I’m getting reasonable success after calibrating properly. My steps per mm were way off and my controllers actually had different microstepa for some reason. So I cleaned that up and did a really good calibration and now I’m cooking ok.
Are people switching away from the MXL belting?
Yes, AIUI, the MXL belting was an early decision based on trying to achieve the $300 price point — it’s intended for timing, not power transmission, and also has backlash issues. GT2 isn’t much more expensive, and it has a tooth profile intended to eliminate backlash, and it’s a bit stronger — that shouldn’t apply at the cutting forces of an SO1/2, but it does help reduce belt stretch.
Have you considered using MXL Kevlar reinforced belts? That is what we use on our SO2.
MXL Series Dust-Free Kevlar reinforced timing belts
I had to go back and check my order for my upgraded parts last year, but I did go to the GT2 belts and pulleys! So that’s good news.
I’ve tackled calibration and am doing pretty well, barring crashing my machine hard last week and I think bending my Acme screw. Very bummed. Super stupid. There’s a bit of wobble in the Z and it will bind on me from time to time. Anyone had any luck straighten it them?
I will say though…obviously…compared to a Shapeoko one with a Dremel on it, this this is amazing. Accurate and powerful. I was actually able to reuse a lot of things in the end.
Thanks again for all the great advice. I really appreciate it!