I'm stuck, lost in circular logic, need some help with ringing
-
As you can see, I'm getting ringing everywhere in every direction, on flats, curves, etc.
It didn't used to be this way. I've made tons of changes lately, too many to track the variables.
I could use some help just resetting my mental debugging process.
-
Posting your print settings (acceleration, speed...) + your config.g + what printer you have might help
-
@Argo I'm really just looking for the process, like
check your machine, no loose parts, belts aren't too tight, etc.
then the configuration calibration steps, like too high E jerk leads to blah blah blah... and so on
-
Ripples like that could be from too much heat leaving the layers below too molten. A good test for that is printing two of the same object side by side. That usually allows for adequate cooling time even though the speed and temp are the same.
Or it could be a bad pulley bearing or loose bolt. Belt rubbing.
Hard to say whatelse could be going on without more info.
-
@Phaedrux thanks.
-
Does the pitch of the ringing vary with print speed, or not?
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42 yes. It does. However, I overlooked this because I just, JUST cleaned them all, I went back and looked at my idlers and they were all gunked up. I cleaned them and it's back to normal.
-
@dc42 something else I noticed. Curved surfaces have no ringing, it's only long, straight, highly accelerated surfaces.
That leads me to think there is tuning to be done around extrusion.
If I set M566 E20-50 and M572 S0.03 or even lower, the acceleration is drastically reduced and the quality is pretty much good enough.
But that makes it really, REALLY slow. Kinda defeats the purpose of having a high perf printer.
I've also tried large print temp ranges with little influence.
Does any of this sound like any extruder misconfiguration you know of?
-
@gnydick said in I'm stuck, lost in circular logic, need some help with ringing:
@dc42 something else I noticed. Curved surfaces have no ringing, it's only long, straight, highly accelerated surfaces.
That leads me to think there is tuning to be done around extrusion.
If I set M566 E20-50 and M572 S0.03 or even lower, the acceleration is drastically reduced and the quality is pretty much good enough.
But that makes it really, REALLY slow. Kinda defeats the purpose of having a high perf printer.
I've also tried large print temp ranges with little influence.
Does any of this sound like any extruder misconfiguration you know of?
Ringing isn't an extrusion phenomenon. It is a lack of nozzle position precision because of vibrations.
I'm guessing that you are using pressure advance, so when you drop your E jerk down, the side effect is that your XY accel gets way reduced, and that is what causes the change in ringing.
Obviously a mechanical once-over is the best place to start. There is no sense tuning around a mechanical issue.
For tuning, I am having really good luck with SuperSlicer custom gcode. You can use it to change accel, jerk, etc per layer. I'd start with pressure advance and dynamic ringing control disabled. Set a low XY accel, like 600. Then, run a test print (a 3 wall, 25mm cube works) that varies XY jerk from 1 up to 20 mm/s, per layer. Then decide what jerk setting you like best. Fix jerk at that value, and make a new print that varies XY accel from 200 - 4000 mm/s^2.
Once you know these kinematic thresholds for your printer, you can add back in pressure advance and dynamic ringing control if you want. But make sure to keep your E jerk high enough, so that pressure advance isn't crippling your XY accel.
-
My preferred way to beat ringing is to use a lower accel and jerk on the outside walls so that no ringing occurs. I still use a high speed (or at least a speed that maximizes the flow rate) so that on longer sections the speed still picks up and slows down for details and cornering. Then for interior sections accel and jerk are increased to minimize print time.
The print takes the time it takes, it doesn't matter to me how long that is. And as long as your close to the max of your flow rate, it can't go any faster anyway. The important thing is that the print succeeds and looks good.
-
@Phaedrux said in I'm stuck, lost in circular logic, need some help with ringing:
My preferred way to beat ringing is to use a lower accel and jerk on the outside walls so that no ringing occurs. I still use a high speed (or at least a speed that maximizes the flow rate) so that on longer sections the speed still picks up and slows down for details and cornering. Then for interior sections accel and jerk are increased to minimize print time.
The print takes the time it takes, it doesn't matter to me how long that is. And as long as your close to the max of your flow rate, it can't go any faster anyway. The important thing is that the print succeeds and looks good.
Worth mentioning a couple things:
-Fast inner walls that ring can print though a slow, non-ringing outer wall.
-This approach probably works better on a direct drive, because low accelerations with high-ish print speeds cause a lot of flow rate mismatch on bowden drives.
