I have a router sled that I built using linear rails/bearings. It works fine but, I want to automate it. I’ve got an Arduino + TB6600 + Nema23 + lead screw solution. I’ve already installed the grbl library on the Arduino. I had plans of using Lightburn (which I purchased for my laser) to control the motors with will use lead screws to move the router (DeWalt) around over the plank with a spoiler bit in it.
Now I am thinking I will write my own app to use a wand and ‘discover’ the shape and whereabouts of the plank to be flattened and then retract the wand and do the work. I just have to learn grbl.
My question is, since I am going to all of this work, why don’t I automate the Z axis as well. Now I am thinking that I don’t want to use the DeWalt router, but maybe a spindle. It would make it possible to control the speed as well. I will still need to come up with a mechanism to control the altitude of the spindle.
You can see where this is going. Now it is a CNC router with a large (300mm x 700mm) work area. Damn. Now I think I might as well make the area bigger. This will never get done.
Those of you that have done this must have already made some of the mistakes that I am going to make.
Ideas?
Find some of the existing builds to start with and see their mechanisms. I started from the OpenBuilds Ox design. You might want to look at the LowRider from V1 Engineering.
I went big (1m x 1.5m rails) and didn’t think enough about deflection. I changed from doubled 2060 extrusion to C-beam 4080 and it’s still barely stiff enough.
Not clear whether you know this, but GRBL isn’t an app you would write a probe-and-flatten in. It’s an interpreter for a relatively limited set of gcode.
Flattening a surface is something you need to do even if only to get a surface tram to your axes for a reference. I wasn’t satisfied by using CAM to generate gcode for surfacing, and wrote something to do this with minimal extra movement:
Thank you for your quick response Michael.
I looked at the LowRider 4. It’s Y-axis is friction based?
I have been of the understanding that screw driven would only be outperformed by ball-Screw driven.
Thanks for the python script. It is at least a toehold, even if I don’t end up using it.
Is the extruded aluminum that much stronger than the steel linear rails? That is counterintuitive.
As for GRBL, I agree that it is not the answer for this project. It would have been fine if I were just going to use Lightburn and 2 axes. But I am upping the ante.
Thanks for the feedback. I am sure to be getting back to you with questions. 