Ripple/vibration test with StealthChop on.

Ripple/vibration test with StealthChop on. Happy to see it eliminates a LOT of vibration coming from abrupt stepper direction changes (like the inset letters and dimples on the sample in the photos). You’d normally see multiple “shadows” of the letter with normal microstepping due to stepper acting like a spring in a spring+mass system.

You can see the corners aren’t 100% perfect though, due to excess extrusion / blobbing i’ve mentioned before.

Download the pics to inspect them at 100% zoom. The layers are very very neat and orderly, but they look a little weird in the downsized images and thumbs due to moire.

I get the funky corners as well.

https://drive.google.com/file/d/0B4vlYd5QDwgBZUlJWmdnOGZvNzQ/view?usp=sharing

Cool, so i’m not crazy :slight_smile:

Me too, that is why I created the issue on this a month or two ago.

While you’re right in that having the Advance algorithm built in would be very very nice, i still think that’s treating the symptom instead of the disease in this case.

Even without Advance, i think there should be less corner overextrusion than there currently is.

Maybe, but that is the whole point of advance algorithm. Without it, redeem will only extrude how much and when it is told. It won’t turn down the extrusion amount when it gets to a corner or before a corner since it is only interpreting the gcode extrude command (again, that is the whole reason Advance or any sort of code like it, as far as I know).

I’ve noticed that I get a better finish and less blob around corners when I turn print speed down, too, but I can’t seem to run the develop branch of redeem. I printed the test cube at my regular speed (60mm/s outer perimeter, 100mm/s infill) and at a very slow speed (25mm/s outer, 40mm/s infill) and the results are clear. Also, stuttering issue goes away when printing slower, so it’s a confirmation of print buffer/path planner performance issue.