Ryan, if you ever do another revision of the MPCNC, please please please use M4 screws instead of 6/32, they were almost impossible to find, even Amazon and Aliexpress proved difficult, and metric screws should be easy to source no matter where you live…
But thank you for this great design, so far it was a fun building experience
The 6-32 screws and the drag chain finaly arrived last week, so here are some more pictures of my build progress:
[attachment file=72061]
The legs stick through the main plate so that I could lift the entire machine by about 20mm if I ever need some more Z space.
[attachment file=72062]
The bottom_corner and lock_corner parts are solvent welded together witch acetone.
[attachment file=72063]
XY gantry, this was a bit difficult to get square, it’s still 1mm off at 300mm distance…
[attachment file=72064]
Z axis.
[attachment file=72065]
The drag chain for the Z axis and spindle/laser cables.
[attachment file=72066]
Drag chain for the Y axis. I use 10mm steel reinforced GT2 belt, I know it’s not necessary but it’s what I had lying around…
[attachment file=72067]
Since the spindle tended to get very hot on my old cnc I decided to print a replacement fan that should keep it cooler. It holds up to 10’000 RPM and moves a lot of air.
I hope to finish the build this week, the only things left are the electronics and the calibration.
I didn’t even think about that, the additional lift should make it a bit lighter ?
And yeah, finished is the wrong word, I hope to get it running so that I can cut the parts for the electronics enclosure.
After that there are some small things I would like to make, and since I live in an apartment I should build an enclosure for the entire MPCNC to reduce the noise a bit…
Have you seen the videos of quadcopters painting fences? I think there’s still a lot physics that aren’t going to be modeled well enough to measured fast enough to make that work… But it would be fun to try! It might be more possible with something that gives you less force back, like extruding paint, and in an environment with accurate measurements, like fixed IR cameras and fiducials on the quad. Maybe lasers?
Since the quadcopter moves by tilting, the spindle would have to be mounted on a gyro. As an added bonus this makes it into a six axis machine
And I don’t remember the name of that router that is a mix between cnc and handheld… It uses a camera, markers on the workpiece and a display to guide the user in moving it, and it has an inch or so in the x and y axis that it can correct if the user pushes it of the path. The quad would need a system like that to correct for turbulences…
I did a quick square calibration and it was only off by less than 0.5mm on a 500mm diagonal.
I have different stepper drivers for the X axis than the rest of the axis, and even though they are all A4988 drivers, the current sensing resistor seems to be different. On the X axis I have set the potentiometer to 0.75V and the drivers get a bit hot, on the Y and Z axis I have it set to 1.2V and they barely get warm. All motors stay cold to the touch.
And here’s the crown test:
[attachment file=“crown test.jpg”]
I didn’t have any paper at home, so the whiteboard and marker had to do
So it turns out that the steel reinforced PU 10mm GT2 belt is too strong, or rather not flexible/bendable enough.
To get it tight enough for it to run nice over the pulley I have to tighten the cable ties so much that the force on the motors is too much and they start to lose steps when going too fast or when encountering a small obstacle.
I ordered some normal glas fiber reinforced rubber 6mm GT2 belt which should fix this problem…
what’s the specs of your spindle? it looks like a generic Chinese unit which i am looking at getting. I’m just not sure how many watts and what controller.
