Prusa Z axis, the best AND the simplest BUT ? by MKSA

Last crawled date: 3 years, 1 month ago

NOTE: This is superseded by https://www.thingiverse.com/thing:2594693 which gets rid of the crappy Geeetech sheet metal parts and the flimsy way it holds the bearing/nuts.
Owners of a Prusa Geeetech Al who have carried out all the mod I published here, mainly the Z axis, ( http://www.thingiverse.com/thing:1830451 ) are the proud owners of the best Prusa, no wobble, noise, vibration, fast, accurate AND simple.
See my other things (bearings, filament guide ...)
Here a video of the printing of a 2cm test cube in ABS at 180mm/s. Quality is still OK but I usually print at 100- 120mm/s to get good results as there is usually no need to go slower.
http://vid1053.photobucket.com/albums/s467/MKSA/20170113_180807_zpsobkt9o31.mp4
I was wondering if I could make the Z axis even simpler by getting rid of the Z smooth rods.
Yes indeed it is possible although I reckon a few issues and limitations in the case of my Prusa.
I was quite surprised by the quality of the TR8 screws, straightness and surface finish and I was wondering if I could use them not only to move but to guide the X carriage and therefore remove the Z smooth rods (OK,not super high quality but suitable for simple, low accuracy FFM 3D printers depositing blobs of molten plastic).
So I removed the Geeetech Al Prusa smooth rods and guides, replaced the TR8 by longer ones, added homemade IGUS bearings in the lower plates, replaced the brass nuts by a housing containing the brass nut at the bottom, an IGUS bearing at the top.
Does it work ? Yes indeed BUT:
The section of the TR8 core is about 6mm thus quite thin. Not much radial forces applied to them, static force due to the gravity center of the X assembly not in the plane formed by the two TR8 plus dynamic force due to the accel/decel X carriage.
Note the moving mass in my case is about 260g. So it is fine. The original MK8 plus carriage (close to 600g) would have certainly been an issue.
The real issue though is that my Igus bearings are a very close fit and if the TR8 are not perfectly aligned, they will require far more torque. The Prusa frame makes it impossible to adjust and keep this //. I had to increase the current close to the motor limit.
I could have devised some adjustment mechanism for // but not worth it.
I also tried to replace the brass nut and bushings by two Igus Tr8 nuts I made. Works too, additional benefit, no backlash and not the slightest play either but again need far more torque.
Now, if this is not really suitable for a Prusa, it shows that for an other frame, a sturdy box frame, one could use three TR8 or better TR10 (to be scaled to the machine size), to design a simple Z table for a CoreXY, H bot ... machine.The key is proper // of the TR8s to avoid excessive friction and binding. This requires mechanisms to be able to adjust and keep the TR8s//. Use cast iron, granite ? Just kidding ! The reward would be a simple no play, no backlash, vibration free system. No need for messy lubrication too !
The ABS and PETG test parts have been made at 100mm/s 0.2mm layer. (nozzle .4)
PS: Ball lead screws would of course even be better but overkill for such simple machines. unless you can get them for cheap ! They would even "straighten" the Prusa frame without binding :)