Building a new printer - Karbon Kossel PRO

[Generic Default]

I'm using the same aluminum plates, some good stuff for the money!

I think your build plate stack is a bit too much, maybe just do

-------------------------1/8" PEI sheet--------------------------
-------------high temp unicorn blood adhesive--------------
------------5/16 MIC-6 aluminum heat spreader------------
---------------------Rev 3 Onyx on 24V------------------------
-----------------------borosilicate glass------------------------

The glass would insulate the bottom of the printer from the aluminum. The aluminum should be in direct contact with the heater and the PEI, the glass will slow down heat up times a lot and it won't add anything useful if it is on top.

I found that the weight of the aluminum plate is enough to prevent sliding. Not sure why you need the glass since plate is so flat already.

Also, if you go with the double sided adhesive most of us are using, make sure you get extra and be careful putting it down. It likes to ball up a lot. Some parts of my plate are a couple of thousandths higher because of the glue stacking. Still crazy flat though!

I considered using red permanent thread lock to bond the PEI to the aluminum plate, but I don't know the operating temperature or bond strength and the experiment would be expensive if I messed it up.

EDIT:

I just realized that unicorn blood contains short lived radioisotopes and provides its own heat source. No need for the heater I guess.

[doctorgonzo]

You only need .04-.03" PEI, 1/8" is pretty overkill, and will probably promote a marginally higher heat gradient.

Thanks Nylocke, I read some of the Amazon reviews and folks are saying the thicker stuff is nicer to work with. I'll try the .030 and maybe get a piece of .060 as well.

Generic, I'd like to be able to flip the glass/adhesive/PEI sandwich, following Mhackney's example. PEI on one side, kapton on the other seems like a pretty catch-all print bed.

[JFettig]

I have only tried .030 - works great. I have a mcmaster sheet of .040 for a future printer.

All I see at one of the places I work the Stratasys uses .030 or less

[doctorgonzo]

My Alligator board shipped on Friday!

Time to get serious about the mechanics on this thing. The only real parts remaining are end stop switch mounts (2mm carbon fiber) and the printed parts for the linear skates. The skate parts clamp and tension the belts, mount the ball bearings for the mag arms and provide a location for the end stop thumb screws. Time to get busy in Rhino.

I'm waiting on a LCD panel, because I'd like to use a color touch screen, or at least a VIKI with the Alligator. I may end up hacking a cheap smart phone for this, not sure yet.

Chris

[doctorgonzo]

OH $#!% it's here!

[img]http://i.imgur.com/7VJbHi5.jpg[/img]

[doctorgonzo]

Alright, I got end stops wired and mounted. Made some 2mm carbon mounts for 'em:

[img]http://i.imgur.com/9aaCuuO.jpg[/img]

I have an extra set if anybody wants them!

I also finished the design on this printed doodad:

[img]http://i.imgur.com/AqvL4e4.jpg[/img]

Here's the whole shebang, sorry for the pic. Hard to get the lighting right, hold the printer and the camera, and maintain a high level of awesomeness simultaneously.

[img]http://i.imgur.com/hoI7Wsa.jpg[/img]

The method for adjusting belt tension on the Kossel is to loosen the FRAME and move it up or down. I think a frame should be as static a part as possible, so...I've been thinking of ways to terminate the belt and make it adjustable for tension, and this is what I came up with. The M5 screw holding the pulley goes in to a captured nut, so you can loosen it, turn the pulley to tighten the belt, then tighten the screw back up. Seems to work pretty well.

Chris

[TFMike]

Have you seen these? https://www.youmagine.com/designs/varia ... a-carriage

https://www.youmagine.com/designs/kosse ... arriage--2

[Generic Default]

I really like what you did with the belt tensioner! I was thinking about doing the same but with an elliptical printed piece instead. The pulley could have a slot cut into the middle so the belt wraps around once with the teeth engaged, but the end is bent into the slot so it can't come out. As long as you keep the belts tight and the backlash nonexistent, your prints are pretty much guaranteed to be good.

[doctorgonzo]

The pulley could have a slot cut into the middle

Genius!

Now, how to cut the slot? I have CNC...make a fixture and use a tiny end mill? Go to town with a cutoff wheel in a dremel? Hmm. Perhaps this is a situation that calls for a hack saw.

It holds fine as is, as the tension of the belt keeps the teeth engaged for about half the circumference of the pulley. I would like the engagement to be more "positive." Something about using belt tension to hold belt tension is elegant, but iffy.

It's all in beta until it's done, and it's never done. This design has evolved a dozen times in my head, on the screen, on the machine. I'm up for whatever improvements make sense.

Chris

[Generic Default]

Two more things I checked on my own printer recently;

1) Measure the distance in mm between outsides of the chrome ball bearings with calipers. Then subtract 9.525mm (assuming 0.375 balls). This distance should be within about 0.05mm for all three carriages and on the effector. I found that the socket cap screws don't center the ball perfectly since it's a headed hexagon indentation, not a 3 point one.


2) Put a metal nut on your Z adjustment screws on the carriages, then clamp it down a bit onto the surface. This prevents the screw from drifting because of vibrations.

[doctorgonzo]

Generic,

I have some of the drilled and tapped ball bearings, M4 thread. I will probably swap over to these, as I'm not a fan of the bearing/socket head/JB weld arrangement. The drilled and tapped bearings are slightly bigger than 3/8, I think they're 10mm. Slight adjustment to firmware but no big deal.

Good idea on the jam nut for the end stop adjustment.

Chris

[doctorgonzo]

I got a bad stepper from Sparkfun so first moves are going to have to wait.