Tips to mitigate vertical artifacting Duet 2 Wifi
-
More pic of the stripes. Reminds me of old school A4988 candy strips.
-
For the sake of completeness can you provide your config.g, the results of M122 and M98 P"config.g". Maybe some more details on the printer.
-
@Phaedrux Sure, I posted config.g above, but I can do it again. config.g m122.txt 10/2/2020, 9:23:33 PM M98 P"config.g"
Warning: Heater 0 appears to be over-powered. If left on at full power, its temperature is predicted to reach 265CThe printer is an EXO cube, almost just like this one.
https://www.exoslide.com/exocube-printer -
@brotherchris said in Tips to mitigate vertical artifacting Duet 2 Wifi:
@Phaedrux I thought about crossing the belt's I might try that. Config and slicer profile attached. I will print a ringing test. slicer settiongs.gcode config.g
Sorry I missed those before.
Nothing stands out as obvious in the config.
Which leaves us with mechanics and the minutiae of stepper drivers.
-
@Phaedrux said in Tips to mitigate vertical artifacting Duet 2 Wifi:
minutiae
Totally agree, I am pretty used to troubleshooting these types of things. I have seen this before, but never on a 2660 driver. This post was mainly to see if any other Duet users were seeing this. Great discussion by the way, enjoying it.
-
You might find this interesting: https://duet3d.dozuki.com/Wiki/Tuning_Stepper_Motor_Drivers
Not sure how applicable or helpful it would be in this case though.
-
@fcwilt Thanks for the reply, Left is normal, right is super tight belts.
-
@Phaedrux Ringing test.
-
Might be able to reduce the ringing a bit with a drop of jerk from 1200 to 700.
Hard to say if that would be enough to fix the benchy artifacts.
-
@brotherchris said in Tips to mitigate vertical artifacting Duet 2 Wifi:
@fcwilt Thanks for the reply, Left is normal, right is super tight belts.
I wish I could remember exactly what I did to fix the problem. It was awhile ago when I had a CoreXY of my own design.
Frederick
-
i switched to 0.9 degrees stepper and 16 tooth pulleys.
-
Latest print, the 0.9 degree steppers are my last resort. I think I have exhausted most of my other ideas. I will switch them out today and report back.
-
Sorry, kinda off topic, but just put 2+2 together and realised @brotherchris was Chris Riley. Love your videos, keep up the fab work
-
@NexxCat lol, thank you!
-
Yeah definitely looking like VFA. Upping your natural steps per mm on XY would be a good next step.
-
The 0.9 motors might have helped a bit, but I think this is as good as it's going to get. 1.8 on the right 0.9 on the left.
-
Did you end up playing with a range of jerk (instantaneous speed change) values? I've found that certain values of jerk are much worse than others for artifacts like this. Obviously, the amount of acceleration plays a role, too.
Try a test of ranges of jerk values with a range of acceleration values, too. I almost forgot to mention feedrate. Try a combination of different jerk, accel, and feedrates.
dc42 has on his list to implement some alternative to "jerk" (not real jerk, but you know what we mean) similar to junction deviation. After finally learning what junction deviation actually is and how exactly it works -- I'm convinced that I would like to try it! It might alleviate problems like this: weird motion on curves.
-
@bot I have made a few changes, my settings are pretty low at the moment, but I could setup a test and try different ranges.
-
@brotherchris it's not exactly putting them low that makes the best results -- it's just finding the right combination. For me, for example, 4, 5, 6, and 7 mm/s jerk produces worse results than 8 mm/s. In some tests. It's hard to figure out, but sometimes you find a combination that works magically. Heh. It's my method for now: guess and check.
-
@bot lol, guess and check is my thing! Testing now, I will see if I can find a combo. Thanks for the tips.