I’m not ready to make the jump to Klipper just yet. I want to iron out hardware issues first.

I have a couple important takeaways. First, build the damn thing to be actually rigid out of the box. I thought I could keep it cheap with only 2020 extrusions, but I’m honestly paying a helluva lot more to add some stiffness to it.

Second, plan for your wiring at any given point to double. I got really tidy with my wiring at first and now it’s an absolute nightmare. I gave up and half of it dangles off the back on the floor. It works but I really could have done a lot better with some better planning. It is truly a shameful sight.

I did add an extra pair of motors to the Z axis, bringing me to a grand total of four. I’m waiting on some stepper drivers to get it going again.

I don’t have any pictures on hand, but could take some later.

I built myself a 400x400x500 CoreXY machine from scratch.

Honestly, my prints are dogshit unless I go slow, and then sometimes they’re dogshit anyway. I have found the entire exercise as one of futility and massive frustration… but I low key love fucking with it. I can’t help myself, I keep building upgrades and improvements for it. I keep sinking time and energy I don’t have (and cash that I do) into this fucking heap of shit and I can’t help but love it to pieces. It’s dumb as fuck and it’s pretty fun to work on.

Honestly the vibes are about the same for a project car. Don’t daily it or you will hate yourself, but if you put it together and tune it right, you’ll have a blast.

Haha! I had a Mini T as my first one. It was a fun little car!

Check out the Tarmo project. I’ve been wanting to put one together for a while.

I realize I’m reviving a super old post here, but I actually did end up figuring out what it was. I posted a little writeup in here if you’re at all curious. You were actually kinda close.

Because I hate when people ask questions and never go back to update their original post when they solve the issue, I’m gonna post an update here and now.

The fix was actually surprising but simple. I ended up moving the second Z axis lead screw carriage mount a bit further towards the motor. That is, the piece that supports the build plate, where it connects to the leadscrew.

Further diagnostics ended up showing me that it was binding, and one of the problems I had later on was that it was binding but only when going up. So I ended up having prints that got progressively more and more stretched out along the z axis.

Moving it to be further in line and loosening a few mount screws to allow the nut to come back to vertical got me to where I needed to be. Hope this helps!

The print before this one, the blobbing/shreds were tall enough that one grabbed the BLTouch probe and bent it a little. I was pretty mad about that tbh.

I’ll double check my connectors. That’s a solid shout actually.

I think I’ll give that a swing. Thanks for your input!

I think you might be onto something but the other direction. I think it’s not moving quite enough per layer.

I haven’t run a temp tower. I’m not fully convinced that it won’t damage itself so I’ve been trying to think through it before trying another print.

Yep! It actually extrudes just fine. I’m starting to wonder if maybe my z axis e-steps are just slightly off and it’s enough to compound over the course of the print. It seems to hit a certain spot and start grinding itself into the print, which suggests that it might not be moving up enough per layer.

I’d be surprised if it was too cold, but I’ve heard stranger. This is PETG at 250. I’ve actually pulled the nozzle and Bowden tube and pushed a very small Allen key through, so it shouldn’t be clogged.

Z height is calibrated, and I’ve calibrated the e-steps for the z axis. I’ve tried disabling the fade height to see if that was doing something but it didn’t seem to make a difference.

I ran through calibrating the extrusion e-steps. It’s feeding 100mm when I tell it to do so. I haven’t messed with the extrusion multiplier, do you have any information about that?

Yep. I calibrated all the e-steps too. Not just the extruder, but the the movement e-steps as well.

If you’re concerned about durability, you could potentially add in some bolts to use as a slip-in rigid spine to further reinforce the design. I’m thinking something similar to how concrete is tensioned to provide rigidity and strength.

I’m spitballing ideas to see if I can spark something workable for you.

Hm. That’s all pretty good advice in general, thank you! Though I’m definitely not afraid to break stuff. I think that’s generally how I learn.

The current motherboard in my printer is almost 100% functional except for one specific module that I completely fried (and as a result, my parts fans run at 100% all the time). I may have shorted a couple wires as I was working on some extensions. Whoops! Though as a result of that, I have a much better handle on how wiring works on these things and it’s a lot simpler than I thought originally. The sparks were fun though!

Honestly I sorta wonder if I should document my build process and post it to Lemmy. It might be cool to have a “Here’s how to convert” guide.

Uh… Whoops, sorry, forgot which timeline I landed in. It gets a little tricky when you start hopping between 'em, y’know?

One option to consider, and one that solved corner warp for me, is adding in a draft shield to your slices. I add in a 10mm draft shield and have yet to have any issues. Ymmv, obviously, but it’s another option to try.

I ended up getting the 410x410 for my Ender 3 a few months ago since I built it out with the Ender Extender kit. I also like it, and the only times I have problems with adhesion is when my filament is wet.

I’m actually making the swap to full linear too. I have the Ender Extender XL kit, and I’m not loving the way the rollers work with it. I printed a dual Z setup but it’s been fiddly. I’m hoping that the linear setup will get me where I want to be.

Are you, by chance, a software developer? It’s the only other context I’ve heard for unit tests, and I would probably have taken a slightly different approach to it.

If I look at the cavity, I see three sections I could divide it up into. The left side with your tees, the middle section, and the far right with the curves. I would likely have modeled each of those little components, printed a section that’s a few layers high to test the fit, and then once I was happy, I’d have joined them all together. Why? Because each failure is smaller, and I can iterate on my design faster.

This isn’t me saying either approach is correct. Just offering an alternative perspective to consider