GT2 31-tooth idler

[orcinus]

I've been fighting vibration and resonance lately (last thing remaining to iron out in my efforts towards "the perfect print", whatever that is) so one of the things i've been overhauling are the belts.

I've switched to GT2 belts (no improvement re: vibration, in fact, it's a bit worse, but the accuracy and repeatability are up) and got annoyed at the belts rubbing and clipping against the smooth idler lips. I've tried nophead's flipped-belts-over-idlers trick, but didn't like the idea of having my belts twisted (plus, the twist was rubbing against the Y plate and the gantry) so i've made some GT2 profile toothy idlers to replace them.



http://www.thingiverse.com/thing:96324

They're a direct replacement for the original idlers, retaining lip and everything - just drop the two bearings and the washer in and mount in place of the old one.

[kbob]

Now you have a chicken and egg problem. To make a perfect print, you need that idler. To print that idler, you need to be able to make perfect prints.

[nixtacy]

I've been fighting vibration and resonance lately....

Me too. I run at 100 to 130 mm/sec print speeds and was getting horrible 'shadowing' at corners.
I added the following to the end of my start g-code in Slic3r, and it's made a wonderful difference.
The *only* problem so far is that prints take considerably longer and the estimated time in slic3r is way off.
I haven't played with the numbers much beyond this, but so far, I love it.
The printer is far smoother, quieter, and pleasant to be around.
I think my prints are nicer too.

Code: Select all

M205 X30                    ; max XY Jerk speed
M201 Y350                  ; max Y acceleration
M201 X350                  ; max X acceleration

[orcinus]

Now you have a chicken and egg problem. To make a perfect print, you need that idler. To print that idler, you need to be able to make perfect prints.

Hahahah, nah, i can print the idler fine. And lots of other fine-detailed stuff.
The only issues i'm having currently are the forementioned vibration (i.e. "rippling") and uneven layers with some filaments. Still haven't switched to thick ACME rods either (using the 6.35mm ones).

Me too. I run at 100 to 130 mm/sec print speeds and was getting horrible 'shadowing' at corners.
I added the following to the end of my start g-code in Slic3r, and it's made a wonderful difference.
The *only* problem so far is that prints take considerably longer and the estimated time in slic3r is way off.

That works, but depends on the kind of prints you do.
Low acceleration and jerk will make your surface detail "go soft". Also, it's bound to cause blobbing at sharp corners.

It's a kludge more than a fix.

I've managed to get the vibration down to manageable levels, at accelerations of 1000-1500 and jerk of 15-20.
I've isolated all the screws with rubber washers, switched to thicker idler and motor plates, replaced the X stepper standoffs with a more rigid, printed mount, padded all the plates with adhesive foam and put feet with thick foam and cork padding on my printer. Also, switched from 24V to 12V supply for stepper drivers and turned on low current microstepping on, isolated the extruder from the gantry with foam and the hotend mounting plate from the extruder and a bunch of other tweaks.

So far, the biggest impact was from an unexpected direction - i've tightened the V-wheels to the point where it's hard to move the gantry and Y carriage by hand. Steppers don't seem to have problems (and they barely get warm), but it's probably increased the wear on wheels, so i've greased the rail a bit with calcite / graphite grease. It works because it behaves as a friction damper.

I've also mounted the fans on plastic brackets with plenty of rubber washers. One of my fans was slightly out of balance and was inducing constant vibration in the X direction which was visible in the prints as thin, regular, vertically striped texture.