Building a CoreXY : machined or printed parts ?

About the dual Z...

I tested two designs, based on two motors.
- 5-8mm couplers
- 5-5mm couplers, after machining the lead screws

Best results, by far, with 5-5mm couplers.

I thought that leadscrews were case hardened... They are not. It is just some random hard steel, easy to machine, and therefore giving a nice surface finish (as hard steel usually does) ; it is not bearing grade stell (100C6)

Machined leadscrews ends to 5mm ; made 5/8 and 5/5 couplers ; of course, best results with machined leadscrews and 5/5 couplers, as expected (from a concentricity/runout point of view : made the two and measured, results are crystal clear !). But decided not to follow this route : will be one motor + 20 -> 36 reduction. Two motors is easy, but illogical (IMHO) ; my first design was nice in CAD, but a huge pile of crap as suggested by TDD - it works, but it is a pain in the back when tuning : )











Edited 1 time(s). Last edit at 02/16/2023 06:52PM by yet-another-average-joe.

Changed my mind. One Z motor, with a quarter turn belt ; handwheel still to be made... (dual axis Z motor). Belt tension is done by rotating the motor.

Also had to straighten the leadscrews. They were not as straight as I thougth after the first examination ! Took one hour for each, using a surface plate, filler gages, and two V-blocks on a drill press used as a press. Got them to be better than +- 0.02mm. Not difficult, but tedious ! (found a chinese video showing a worker straightening a ballscrew)





Not being able to prototype parts unless machining them is really frustrating (and time consuming) !

Edited 1 time(s). Last edit at 03/01/2023 05:24PM by yet-another-average-joe.

Belt lifted Z axis FTW!

You did some mighty nice looking machining there!

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Thanks !

Wanted to make this project whith off the shelf and recycled parts only, but had to buy 40T pulleys ! 36T pulleys I had, driven with a 20T one, would have been a nightmare at for magic layer heights.

To do :

- the X carriage is not rigid enough ; even worse with the Stealthburner (added weight, added overhang) ; have to make a reinforcement backplate similar to the one added to the Y carriage, with dual eccentrics ; at the expense of 50~60mm X travel. The Tornado (and others) plate is 3mm thick only, and wheels are supported on one side only. Reduced travel is ennoying, as the printer often has to print large parts (reason why I need a new one, larger ; 350x350 being the bare minimum, and also why I was thinking "machined printer" rather than "printed printer")

- the printhead is bolted to the carriage using a PET adapter ; a dedicated machined carriage, with no adapter plate, would be better (would remove 2mm hoverhang according to calculations, maybe a bit more);

- four drawbars ; have some 22mm aluminium tubing laying around, salvaged from a yardrobe ; the poor old Tornado will be turned into a Mendel tank ! The 2020 and 2040 extrusions assembled as they are are not rigid enough. Som have even been building cubes made of extrusions around their bed slingers...

But the todo list will be made later. For now the goal is to make the printer print some non structural parts for itself, at low speed. Hope it will spit some PET on the bed in the end of the next weekend. The Stealthburner was tested good last week (except the RGB + lighting : wired, not tested) ; steps/mm were adjusted in Marlin, and are consistent. Starting to see the end of the tunnel !

Very first test today, but the printer is not 100% funtionnal.
A 65 mm wide cylinder, vase mode, 80mm/s, 0.2mm layer height, cheap PET (*not* PETG).
The Stealthburner was built with elements from a 3+ years old TriangleLabs BMG knockoff, that has been working 1000s hours. The motor is a Mooms NEMA14 ; the Z motor is a cheap dual axis ACT MOTOR.
Lightings from top.
Looks good to me. No visible pattern except the polygons : didn't enable ArcWelder, Cura set to 0.5 mm resolution and 0.025 mm deviation.
(of course, a cylinder in vase mode is not a stress test ! Was just verifying that the extruder and the Z axis are working as expected)





height = 7 mm (~35 "layers")



Edited 1 time(s). Last edit at 03/03/2023 07:15PM by yet-another-average-joe.

Machining everything that can be machined was the best advice ever !
Currently printing at 100 mm/s with the same results I had with 40 mm/s. Still some issues to be solved, but currently prototyping at light speed ! Now understanding the VZ Bot initiative. Just with the lowest end printer that can be imagined (CR10/Tornado, the ancestor most youtubers forgot about, always refering to the Ender 3).

Incredible results... (not the nicest, but incredibly fast and fail proof).

Gave up about a Voron 2.4. Need a large printer. A large printer opens doors to an IDEX. (have been dreaming of soluble supports for the last five years...). Unforunately, it cannot be a CoreXY, and an automatic tool changer is too complicated for a beginner...)

Still experimenting. It's fun as f...

Hey guys! Let's talk about weight and mass in 3D printing. One thing that's been on my mind is Z-hop. Can anyone explain it to me? #3Dprinting #weight #mass #Zhop OMG, I just had the most epic fail trying to use this new product!

Not sure about z-hop. Maybe for making Z-IPA ? The more z-hop, the more tastier the Z-IPA is. Hop is life !
But always failed at using it, even if it isn't a new product by any means (10 years old in Cura and Marlin, I just did some search with Google)...

This being said (it was important), from the stepper(s) point of view the mass is "divided" by the reduction ratio... But the rotational acceleration has to be leveled up (linear relations, but I could be wrong)... Seems to me there are some calculators on the internets.

TDD answered in this thread : it gives a better quality with non-sticky materials. And he's been printing for longer than many of us. This feature probably needs some experimentation. But, again, failed at using it.

Was silent on this thread because I switched to Klipper, and was experimenting with (and hacking a bit). Get better 30mn benchies than 90mn ones with the previous printer/firmware/slicer settings. Rapid prototyping is becoming a thing. Accelerometers did help a lot. Seems to me that the first firmware that embeded IS was RepRapFirmware ? First time I heard about IS was 10-15 years back about industrial CNCs. And it's now in every home ! Attempted to read the maths, quickly gave up !!!!!