I use 3-in-1 oil whenever weird noises creep up.
I’ve been studying in much of my free time as well and I have some results to share. Word around the internet is that the "ringing’ is caused by mass at speed subject to sudden changes in direction.The mass of the stopped axis bounces or oscillates. thawkins’s response to this problem posed on the Printrbot forum made sense to me. I only came across this post today and I had already been chasing my tail with various settings as you’ll see below in results numbered 1-5.
In my tests I used larryj’s 20mm calibration cube. I’m using Cura. I’m convinced I should be using simplify3d, I just can’t bring myself to drop the $$$. I stared out trying to achieve quality at 60mm/s but changed gears 2/3rds the way through. In all the the tests temperatures were constant,layer height was 0.2mm and shell thickness was 0.8mm (2 layers thick), If I didn’t specify jerk or acceleration settings then I’m using Marty’s Marlin defaults.
There is much hideousness in prints 0-4, printed at 60mm/s. So much that’s it’s hard to discern the ringing/waves. In person I can see and feel them and they all suffer the artifact worse than print 5.
I don’t expect anyone to be able to read my chicken scratch (hand writing) or care about the round-about way I came to a reasonable result. I’ll just jump to the best result and share how I achieved that and share some of the things I’ve learned along the way.
This is a close up of #6 above. The best 20mm calibration cube of 11 that I’ve managed to print. Don’t worry, I haven’t written home about it.
In short, I achieved the best results by using 4 top and bottom layers and lowering the speed of the outer shell printing to 15mm/s. Here are the print speed settings I used form the Advanced tab of Cura.
(In my case) The imprint of the ringing/waves seems to be linear to speed. At 60mm/s the imprint may be 10mm wide, at 30mm/s the imprint may be 5mm wide.
I did tests with very taunt and loose (not slack) belts on the Y and X axis and it didn’t create any noticeable results in this artifact on the Wilson 2.
Reducing Accel and Vxy-jerk in the Control/Motion settings of the printer did have a slight affect on the amplitude of the waves but didn’t reduce there overall imprint.
I’m curious if any firmware settings or slicer settings can tell the printer to slow down as it approaches such an axis stop.
I suspect that a reverse movement of an axis would damper this affect and I’m curious if one could print at higher speed it printing right angles of parts 30-60° offset to the cartesian plane. I kinda want to test this but I’m also very bored now of printing this cube.
Hope that helps someone other than myself.