Hey guys, I am hobbyist looking for answers. I am finishing setting up a custom CNC router build that I bought from someone. This machine was designed for ski manufacturing so it has unique dimensions. -custom designed CNC router loosely based on the Ox CNC and openbuilds parts. -working area of approx 84" x 15" x 6" -DUAL rack and pinion drive on the long axis(X), lead screw on the other 2 (Y & Z) -24V / 14.6A Power Supply (Mean Well LRS-350-24) -CNC xPRO V3 Controller Stepper Driver (matte finish) When I purchased the machine it was basically just parts with a rough idea of how it goes together(there was a solidworks file but I wasn't able to open it). Long story short everything is together and it runs, just not properly on the electronics front. I have done a little bit of research on here but there is SO MUCH info, its hard to nail down exactly what I am looking for so that is why I have started this thread. I think most of it is properly setting up the GRBL settings, but I definitely need help getting it right. -CNC xPRO V3 Controller Stepper Driver (matte finish - marked 2016 on bottom near power input - 4516 marked on bottom under Z Motor input) •I have flashed with GRBL version 1.1g •When connected to PC(Windows 10 Laptop) and power(24V / 14.6A Power Supply - Mean Well LRS-350-24) all of the LEDs for the motor controls are red when first powered on. I am still able to jog and run G-CODE. When running LEDs flash red and green. When LEDs are red and nothing is moving the motors are humming. After jogging the LEDs can be either all green, all red, a combo of both, or off completely depending on where the motor stops. • $4 is set to DISABLE •uStep setup for 8uSteps/Step(recommended setting) •When running G-CODE all seems to operating well. After a few lines of code my X-Axis stops working and might pick up a couple of other lines, but after a couple of seconds its just the Y & Z that work properly. •This build was designed for dual motors on the X-Axis. I have de-soldered the default Y-Axis control and soldered the X-Axis control. Action seems good as both motor turn when asked to. It does seem that one motor kicks on ever so slightly sooner than the other one so its not a super smooth motion. I have since taken the gear off the motor and unplugged the cable to the motor for continued testing. •All voltage currents for motors are at the preset factory settings, about halfway turned up. @Peter Van Der Walt I see your expertise around this forum regarding the Spark xPRO V3. Do you have any advice/input as far as how this board is acting? The other issue is figuring out the proper settings for the X-Axis. I am not quite sure how what the proper steps/mm should be. The specs on the X-Axis rack and pinion. McMaster-Carr McMaster-Carr The specs for the Y-Axis lead screw. 8mm Metric Acme Lead Screw That sums it all up basically for now. I greatly appreciate all the help!
The formula you need for figuring out the rack and pinion is: steps_per_mm = (steps_per_revolution*microsteps)/mm_per_rev So I believe that gear travels .75 inches, or 19.05 mm per revolution. Please correct me if I am wrong. Therefore, the formula (200 x 8 microsteps)/19.05 = 83.989501312336 steps per mm. The Openbuilds 8mm screw should be about 200 steps/mm. I think they recommend 199.1, though for their machines. This formula is from the GRBL wiki page which is a phenomenal resource for all questions GRBL. After changing the value, if it needs to be changed, further fine tune the calibration like in this low budget 2:30 video. I don't try to make money from videos so I don't ramble on for 15 minutes explaining something that can be explained in 2:30. We probably all get enough 3 day trainings that could be taught in one at work, right.
Since I was able to jog the machine, I attached a mechanical pencil to the Z axis and laid down some lines. At the default setting 200 steps/mm when called to travel 1mm it actually traveled closer to 5mm on the X axis. Y axis is a lead screw and it traveled pretty much 1mm when called to. So I estimated it should be more like 62.5 steps/mm and it appears to be close as the machine now travels approximately 1mm when called to. I just wanted to see if there was an exact way to arrive at that number so I am not guessing.
0.75" is the pitch diameter. Distance traveled is based on pitch circumference which would be 0.75" x (pi) = 2.3562" or 59.847mm per revolution. You can handle the rest. BTW, is it direct drive or does it have belt reduction?
Do you have a meter stick so you can have it travel further during calibration? It will be more accurate.
I have a meter stick. I will make some adjustments and give it another shot. My biggest issues are the control board and making sure its operating properly. I'd really like to figure out how to make that X axis stop working after g-code is started. It is direct drive. I should take some pictures of the setup and post them here.
Upgrade to a BlackBox ? https://docs.openbuilds.com/blackbox for all the info - its got really nice drivers onboard (much better than the DRV8825s)
Thanks for all the help so far guys! (200x8)/59.847=26.7348 steps per mm for the gear driven X Axis.This value when combined with 199.1 steps per mm for the Y and Z axis yields good results. Test plots with pencil show that everything appears to be proportional. No steps are skipping when running g-code now. Is 26.7348 too low of a number? At this rate the motor works, but the feed isn't that smooth or fast(regardless of what I set the acceleration rate to). I tried changing the microstepping to 32 and recalculated the number. @ 32 microsteps the new value is 106.939 steps per mm. When jogging the motor action seemed greatly improved. However when running a test plot the proportions were skewed. Does the microstepping need to match the other axis to be proportional. Is this right?
I would stick with 8 microsteps. How fast is your feed speed (not acceleration rate)? I once ran a project at 127mm/min because I forgot to change the default value in the Fengrave software. It was soooooooooo sloooooooow and the machine sounded like it was going to vibrate apart. If you are cutting wood, I have found that 2500 mm/ minute with the depth of cut being 1/2 the diameter of the endmill and the spindle at 18000 rpms will easily cut any wood.
Agreed, remember, increasing microstepping, decreases torque - 1/8th is a good compromise for CNC applications
I have a question about running the OpenBuilds_CONTROL software on a raspberry pi. I was trying the no-compile method but I think I am having troubles with the NPM install and I think that is preventing me from running the program. This is what happens when installing NPM. I am not an expert with Linux so any help would be greatly appreciated. I have a windows 10 laptop but would prefer to use the Pi for the main control of the system if possible. I was able to run a g-code file via Chiipeppr and there didn't seem to be any issues.
We've found the Pi (last tested with a 3+, haven't had a 4 across my desk yet) choked up on larger files, which really made it a "we won't go there" And then of course from a support perspective, we cannot build installers in the cloud (appveyor/travisci setup we use for win, linux, mac) to do Pi too, and the support load of guiding users through linux is immense too. Also, used PCs and Laptops on craigslist can be had for free (; So therefore the big warning at the top of the Wiki page We rely on fluent community members to troubleshoot issues (as mentioned, only try this if you are familiar with node, npm, linux, etc) and if the Wiki needs updating, let us know. A little handholding is OK, but with anything major, that's really up to you for wanting to use an unsupported, slow platform - things change (updated node version, update npm dependencies, and updated linux packages) Read the errors, and try the suggestions (npm install jscs @latest) - worth a shot, the top screenshot says one of the dependencies has a dependency on that package
I upgraded my OS software on my Raspberry Pi 4(4gb) to iRaspbian and the no compile method for the Open Builds Control software boots up with no issues. I was able to job my machine without issue and load and execute gcodes. Now there seems to be an issue with my Spark Concepts v3 board. The board won't power the motors if the USB is plugged into a computer. If just power is connected to the board the motors are powered on. Then when the USB is connected they cut out and machine cannot be moved via control. I have tried reflashing the board and using it with my windows 10 laptop but I can't seem to figure out what is causing this.
Check that $4=1 has been set. xPro and BlackBox both need that (has drivers that use an inverted enable signal)
