Cheers. The bed is 300x300, but i've sized it to use as many 500mm extrusions as possible to save on cutting. That considered, it could probably manage 400x400, but i thought that was overkill, and an easy change at a later date if needed.

I'll take a look at some existing printers, good shout. A chap on IRC mentioned a upward facing bolt with a spring, and a nut capturing the build platform, which i quite like...

 

I had started my belt path exactly like yours, with the idlers mounted on a gantry plate. I found the my frame was excessively not rigid, imagine standing over your bot, grab it like a steering wheel, you could almost make a turn with how much torsion there was. My solution to that was to add 3 additional cross-members, one under my Y axis gantry sliders on each side and an additional on the floor. Made a good place to mount the idlers. you can remove that plate and one idler from your part count while stiffing up your frame. Eric was talking about bolting his to a wall. At this Z height, serious reinforcing will need to be done. I would like to further reinforce my frame, possibly with 4 mill acrylic panels on three sides.

 

I'll take a look at the triple c-bot again. Thanks Adam. I assumed your extrusions under the Y axis gantry were at different Z heights to your lower X-axis extrusions to suit your belt layout.

It's an easy change to just raise the lower left and right Y-axis extrusions, and even to add acrylic panels like you said. Which if you say improves rigidity in torsion, I'll most likely do.

I have a look around for an acrylic panel mount that fits v-slot.

 

My build plate consists of (from top down) a large piece of glass mirror, a 1/4" thick aluminum plate, and a 110v silicone heat pad.

What I ended up doing was drilling three holes in the aluminum plate (two corners and one in the middle, opposite) large enough for an M3 cap screw, for a triple lead screw set up.

Then, I drilled a larger hole half-way through so that the socket cap can fit flush with the build plate surface.

Finally, I used a spring, printed thumbwheel and an M3 nut for fine-tuning the leveling. Unlike some other builds I've seen, the springs I use only help keep tension between the aluminum plate and the thumbwheel, so . This means that the whole build plate (including leveling hardware) sort of rests on top of the moving Z platform. If I need to remove the build plate, I simply lift it off.

I'm printing out some long frame braces right now, but when it finishes I'll try to post some pics...

 

Not a really good shot, but I have the socket cap screw inserted from the top, then two M3 nuts to lock the screw in place. Then the spring and thumbwheel. Inside the thumbwheel is an M3 nut. Then the rest of the socket cap screw just sits inside a hole in the Z-stage mounting bracket.

 

I'm doing the same, only I put the thumb wheel in a slot so it is trapped. I like your setup and going to try it out tonight.

 

Has your bed ever moved during a print? Or do the holes in the mounting brackets have very little clearance to the threads?

 

No, my bed hasn't moved... I'm pretty sure even if the holes in the mounting brackets were large, it still wouldn't move... The main reason is that the thing weighs so much...

 

Good point. Hadn't thought about the hotend hitting it. I'll have a re-think.

I didn't think too much of using 20x40 for rigidity lower down on the frame. I was planning to add corner brackets and possibly printing some brackets for larger 45 degree braces where needed with some of the offcuts I'll have.

 

Has anyone done an eccentric nut mod for all the axes? I found one for the xy axis ends on sheffdog's blog but no such luck on the Z or extruder carriage.

 

That's understandable - adds $40 or so at least at list price... should be possible to design a 3D printed version, though? I'll give that a try and do some testing.

 

With toothed belts, I'd say tight enough to not skip, or have any slack-- some amount of "thud" when you pluck the belt, but not a "twang". I think people tend to over-tighten belts by default-- a loose belt can easily be tightened up to remove skipping, but a too-tight belt can damage the printer.

 

Wow you guys have been posting a lot! I haven't been around much the past two weeks but I just wanted to check in and say that the V2 direct extruder carriage is A-#1 Duke of New C-Bot. The amount the motor is shifted over is perfect and makes loading new filament a snap. The holes in the backplate are great for mounting my endstop and the cooling fan. Assembly was even easier now that the "bowden" portion of the tubing is a screw in instead of needing to be installed before anything else. Every modification a win in my book.

I finally settled on retraction of `M207 S0.9 F3600 Z0.0` and recover at `M208 S0.0 F1500` (0.9mm @60mm/s and 25mm/s recovery). The much much much shorter retraction (0.9mm vs 8mm) more than makes up for the lowered acceleration (1250mm/ss vs 3000mm/ss) required to run with no ringing.

Every print is fantastic and just what I was hoping for when I decided to build a new printer almost 6 months ago. I have 270x260 usable build surface so I can print 4x the quantity I was printing before, and at 75mm/s perimeter speeds vs 60mm/s on my i3. My normal print went from 4 hours with a cooldown / heatup in the middle to just 6 hours to produce 2x items. It makes it a lot easier to churn out parts when you only have to check the printer once to get 4 parts and let it go for 6 hours versus checking the start of every part and clean the build plate every 2 hours. Much thanks to everyone who contributed parts, build ideas, concepts, advice, and most of all the genesis @Carl Feniak!

The only thing I have left to contribute is maybe consider moving the XY steppers away from the body/frame a bit. They rest against the printer which is great for stability but they transfer their vibrations into the frame and even with dampers they can cause a lot of buzzing. I stuffed a piece of 2mm neoprene in between and it quieted them significantly, but if they weren't touching at all I think they'd be even quieter? It's on my todo list to try to modify the brackets and try.

 

To echo @CapnBry, yah, the new direct-drive is awesome. I just printed the below Eiffel Tower: 69 hours, 150 micron, .6mm nozzle, 60mm/sec, 20" tall. The Bowen would have never survived all the retractions, and even if it did, the stringing would have been crazy. It did actually fail right at the first landing: There are a bunch of really small handrails, and I have a feeling that the constant retractions there jammed the nozzle and notched the filament. But I was able to re-slice the model at that point, print out the rest, and glue it together. Did my first OctoPrint timelapse as well. Hopefully get the quality better in the future:

 

Just a quick thought,
Could it be due to the coreXY design? In order to move "straight" you will need to move both the stepper at the same time and even though theres is some micro stepping going on the 1.8 degree step will still be more
pronounced then the smaller ones... Try to measure the length between each ringing and see if it corresponds to you gt2 uplley or try to place the object diagonally in the slicer prio to printing

 

I don't think so.
The accel & jerk settings r typically what causes ringing. If the belts r not tight enough I could see that as causing ringing. I am printing flat pieces with no detail @ 100mm/sec and see no ringing at all. It surely isn't an issue with coreXY.

 

the typical answer for tightening "belts" is 1/2" deflection when pushes with moderate pressure.
Of course these r different but the feel is important.
mine flex about 3/8 to 1/2" with the carriage at one side when i pull them, Anything past that requires more then a gentle grip on the hook I use. I need to find my spring gauge and give u a reading. I have not had any issue with them skipping.

 

this is an uncalibrated spring gauge but a quick check with my scale indicated that it is close enough for government work.

With the carriage to the right (~36cm of belt (12x12 bed)) it takes 8oz of pull to deflect the front belt 10mm and 10oz to deflect the rear belt 10mm.
If I cut the span of the belt to about 20mm (an 8x8 bed) it is about 8oz for 5mm deflection for the front.

that should give you some kind of ref point..

 

Hey, I've been looking at this printer and am probably going to make it soon. I was looking at the spreadsheet and was wondering what the "1mm shim, 5mm ID" was. Also, I was looking at the way the X-Axis moves and it seems quite strange that you would have the stepper motor and belt not connected to the axis itself. If possible, could you upload a video on how it moves. Thank you very much!

 

Not having the stepper on the axis is a strong point of coreXY. Not having the wiring for the stepper moving is a big bonus. Keeping the weight of the stepper of the axis also helps with print quality. One of the biggest advantages of coreXY is the 45 degree rotation on the stepper movement. As far a how it moves, imagine an etch a sketch, turn one side and the print head moves one direction, turn both to get another direction.

Those shims go inside and outside of each wheel and allow the bearing to move while under compression. Each wheel uses three, bolt shim bearing shim bearing shim nut.

 

Here is a video demonstration of a C-bot running.

 

I like your idler brackets, little longer.

 

Hello Truculent, I'm still working on the Z axis eccentric mounts, I'll post up the STL files once I've verified that they work.
The Y axis and Extruder(X axis) are working great so far.

Here is the back plate for the hot end(extruder) . extruder_backplate.stl