Wavy sides, Repetier, original style Rostock

Puzzled newb here. Finally getting respectable prints out of a 2 week old Rostock.

I'm having trouble (or not... perhaps I'm being anal...) with the sides of straight prints. In this case, a 25mm cal cube.

I'm getting very consistent ripples -- for lack of a better word -- about 3 mm apart.

I've seen comments about changing some buffer setting, but that seems to be Marlin related, and this is running Repetier.

Any ideas? Aluminum rods, Traxxas joints...

-Ed

It is not related to delta segments per second, the period of the waves is about 3 mm which is too small for that. I read somewhere that small waves can be caused by pulleys which do not match belt. But I never saw it myself.

I noticed that the wave pattern didn't seem to exist on the bottom few layers... I'm going to print another one today with octagonal infill turned off... I'm thinking perhaps the infill distorted the outer walls.

It was a definite smack in the forehead moment last night as I was brainstorming...

So much for my brainstorming. Wasn't even a light drizzle.

Changed the infill and the wavy edge still persists. Next, I plan on changing the microstepping and seeing if bigger steps changes the size of the waves.

More interesting would be this info:
* what is the period of waves compared to the pitch of your bets/pulleys
* how do the waves change when the printed object is in the center of the bed compared to the position most far away from a tower base and most near to a tower base
* are your belts tightened well
* compare microstep size (mosttly it is 0.0125 mm) to the estimated height of the waves (amplitude); maybe it can be related to microsteping if wave amplitude is comparagle to microstep size (typically 0.0125mm), but I kind of doubt that, form the picture it looks like the wave height is significantly bigger than one microstep ... and if so it cannot be related
* if you print slowly (about 2 cm/s) and your delta segments per second is about 100, you can be sure this is not related to delta segment per second at all)

Lets assume one tower has cariage positioning error and the rest do not. The point is that the tower cariage position errors lead to XY position errors (in your case periodic waves) of almost nothing near tower base, of about half the size in the center of the bed and of about 2 times the size 10% from the far bed edge. So if only one tower has periodic positioning erorrs you can find out which one by comparing wave amplitude of cubes printed few mm away from the bed edge (always take places most far away from a tower (which is in between the other two towers)). Those are the places which highlight XY position error of one tower while minimizing XY position error of the other two towers. Preferably print cubes so that their diagonal is aligned with the line connecting cube center and the tower.

Edit 1 &2: Actually it would be more interesting to print half cubes (cut through cube diagonal). This way you will get a triangular base. Align with tower so that the longest triangle side is perpendicular to the connection with the tower and the triangle points (as an arrow) to the tower base.

Just few ideas what I would consider important to check if I would be in your situaion.

Edited 2 time(s). Last edit at 09/11/2013 07:49AM by hercek.

Looks like I have some cubes to print today. Thanks!