Post by jimustanguitar on Feb 8, 2015 3:24:09 GMT -5
So Glenn has his horizontal banding... I've got vertical banding. Trying a new glossier filament really makes it stand out.
The SeeMe guys had a nice long discussion about it last year: forum.seemecnc.com/viewtopic.php?f=54&t=4203& And the conclusion was that all CNC machines do it because of how stepper motors step. Machines that aren't built out of iron tend to transfer those vibrations into the work. More steps, higher processing power (enabling more segments per move), etc seemed to help slightly.
I agree with Phil. I haven;'t taken the time to run test yet but something Graber said at one of the meetings got me thinking. Before I rebuilt my machine I was printing at around 50 to 55 on my speed. After I rebuilt I slowed it down to 30 and 35, got better prints but that's when the bands appeared.
I talked a little bit to Brian and we are thinking that because of the math certain speeds might have odd results. Kind of like driving on an unbalanced tire. 60mph might be fine but 61mph shakes the heck out of your car. But like I said, I haven't run any test yet.
Last Edit: Feb 8, 2015 11:52:12 GMT -5 by BeeAmaker
Some things are meant to be closed. Your mind isn't one of them.
Post by jimustanguitar on Feb 8, 2015 23:52:30 GMT -5
Sliced at 40mm/s. The 4 stripes are 100%, 150, 200, and 250 - so that turns into 40, 60, 80, and 100 mm/s.
One thing that I noticed as I ran the print was that my hot-end wasn't keeping up, so I'll have to re-tune my PID's to print faster and faster. I think that's part of why it's a more matte finish the further right you go.
Last Edit: Feb 8, 2015 23:53:15 GMT -5 by jimustanguitar
Post by bvandiepenbos on Feb 10, 2015 17:56:12 GMT -5
I have found that adjusting speed at LCD does not have same results as changing your sliced speed. Probably best to test this by adjusting speeds of slice and leave LCD at 100% for all.
Post by jimustanguitar on Feb 12, 2015 16:40:17 GMT -5
So what if this banding is the result of pulsed extrusion? The extruder does one 1 move per line of GCode, with a lot of little tiny pauses between each instruction... Could it be that simple?
Maybe changing the microstepping on the extruder or switching to 0.9 degree motors on the extruder would show a change in this behavior?