The Hexagon hotend tends to release excess plastic at speed in corners sometimes. The fix at the moment seems to be switching to an E3Dv6 barrel, with either the stock Lulzbot thermistor or the E3Dv6 thermistor. The other cause of corner wierdness can be lifting. If you have a picture of the corners you want to post I might be able to tell you where to focus your efforts.
Thank you!
The Z axis upgrade helps a little bit, but for the cost and amount of effort it takes to install, it’s probably not worth it for most machines. A Hardened Z rod and higher tolerance Z bearings is probably almost as good there.
The Aluminum bed to main frame connectors are the 4 pieces that join the Y axis bed frame to the Main Taz frame. The plastic ones can flex a little, which allows the bed to go out of allignment just slightly on Y travel moves (we’re talking 0.01mm, but it still counts). The metal ones are more rigid and don’t allow that movement. The top plate is the more important part there.
It is a little on the big side heh. It’s not all that much heavier than the stock one if you include the stock fans on the stock one. I do need to re-engineer a couple pieces of it eventually though. It’s on the list heh.
Thank you! That is a great place to start. I just ran mine at 150mm/s last night for most thingsm and 120mm/s for infill. It did pretty well. Hoping to crank it up a little more tonight
From your pictures there, I woulld bet you currently have it set to print a solid layer every 5-7 layers. That layer is printing at 100% infil and ending up wider than the other layers. Which is indicitive that you might be overextruding slightly.
Canceled the print earlier as it was simply just a test. Top half of the print is at 2000 Acceleration and the bottom half is at 1100. You can see where I changed it. Speed was originally set to 200mm/s but then I upped it in via LCD to 999 (percent?)
Obviously I’m not actually trying to acheive these speeds, but its a good example of the corners as they are not much better with just the standard 500 acceleration and 150 mm/s
I’ll have to take a look at your settings and play around with it the next few days.
(look at the smoothness on those straight runs though!). The one bad side was the side facing the fan. Seemed like the ABS didnt like the fan when it was going that fast. But honestly the speeds are so fast you pretty much need the fan on full blast. Posted the speed trails on my social media pages if anyone is interested. Just search for Mo3D Printing. Still need to post the pictures of the upgrades to thingiverse.
I did note that you CANNOT print with the standard infill unless its 100% or close. The way the infill works by going back over the line perpendicularly thats previously been printed… lets just say at those speeds even with such a small speed bump it gets a little rough on the printer.
I had adhesion issues with the bed @ 80C. Not sure if it wasn’t hot enough or if I didn’t clean it thoroughly enough before the print. Ended up with super warped corners on the front left and right bottom of the cube.
I’m curious, Have you tweaked your “jerk” settings at all? I didn’t see anything OK them in your profile. What do you have yours set to?
I’ve had great luck printing simple objects with your settings (like a cube). You can still see a bit of a seam in the corners though. Now I’m just trying to dial in more complex prints.
Ive played with them a little, but I hadn’t noticed that adjusting them did all that much at the highest speeds, because the extruder was the limiting factor. I have a volcano extruder piece to add to my new E3Dv6 extruder body as soon as I figure out a way to fit it on the machine so I plan on trying again at that point.
Sorry I should have clarified. I’m more referring to your Standart print settings of 90mm/s to 150mm/s. At this point I’m trying to just achieve a realistic print speed for quality printed parts. There tends to always be something subpar every time I attempt to print faster like 90mm/s of anything beyond a calibration cube. Maybe I should just continue to print slow and enjoy the great print quality.
You will hit a lot of new limits with this print speeds when you leave the test cube area:
.) You may see the extruder limit more visualy at “real” prints - don’t go beyond 9-10mm³/s for a stock hexagon hotend.
.) Jerk will limit speed at nearly each direction change, which makes printing arcs very jerky.
.) Last but not least, Marlin has a hard time to calculate all new movements when you approach more than a specific number of line points /second. At some point, this produces a buffer underrun which also results in jerky movements because Marlin tries to reduce the speed to nearly 0 until the buffer is refilled.
Even at 60mm/s Marlin stalls at arcs of some models if you don’t reduce the model resolution. So I think speeds up to 150mm/s are realy limited to technical parts with long straight lines.
Interesting. I didn’t even think about Marlin and software being a limiting factor.
That’s really good information to know.
So I guess on parts that we care about having a some what decent surface finish on (especially on curved surfaces) We are limited to lower speeds due to software?
It’s an interesting point, because when I started printing parts with a large number of arcs. Even at 90 mm/s I was seeing an irrational number of blobs. And it’s not my extruder, its dialed in for this filament, anything lower and I’d see gaps between top layer surface and wall lines. (It could always be the hexagonal hot end, it does seem to struggle with arcs and blobs)
I might try lowering the acceleration back down to 500… But at some point it’s more of “what’s the point of upping print speed” since acceleration will always limit it
So realistically it looks like we are still limited to close to stock lulzbot print speeds, even with modification? Still nothing to complain about as the modifications have significantly improved print quality.
It’s the power of the CPU used in all printer boards that limits the rate at which steppers can be moved and at which rate new points can be calculated. That’s why Marlin may shift to another hardware in the future…
But other limits like jerk limit which limits arc point count are simnple physic and no fast CPU can solve that problem…
