Vcore 4 endless ringing

Hey guys I’m desperately trying to get rid of ringing. I have the 500mm vcore 4 hybrid and can’t get rid of ringing on my benchy. I have spend hours and hours tuning the belts and making sure everything is aligned, but something just doesn’t seem to work. I’m using prusament pla and have turned the speeds down to 2000mm/^2 external perimeters and 50mm/s outside walls. Btw this is only using the core xy kinematics, as I haven’t added the hybrid stuff yet. I also attached my input shaper graphs as well as belt tension graphs. I’m desperate so any advice is welcomed!!!
No description
No description
No description
No description
No description
48 Replies
VisualTech48
VisualTech484w ago
Did you actually put it into your printer cfg?
cm
cm4w ago
Yeah, here it is.
D_Jespersen (DJ_Designs)
always use mzv zv is not effective
No description
cm
cm4w ago
oh really? thanks, let me try that
cm
cm4w ago
So that seemed to help, but still not anywhere near perfect. I printed a bench with the same filament on an old prusa mk3s+ I have laying around and it was wayyyy better quality.
cm
cm4w ago
I even tried using the ei shaper to try to be more dramatic, but still not much better
cm
cm4w ago
And these are the setting I have slowed the printer down to. I feel like this printer shouldn't be having this many ringing issues?
No description
No description
VisualTech48
VisualTech484w ago
I'm wtill wondering if it uses it. Can you share klipper log?
Prav
Prav4w ago
if you run internal perimeters much faster than internal, the artifacts from them can show outside as well. You are going to switch to hybrid anyway so there's not much sense in trying to tune corexy print quality. because there's a bunch of stuff that make the surface quality shitty that's not directly related to input shaping. eg. if PA is off it can manifest in similar way or magnify visible ringing. in your place, i'd try to find a reason for that 100hz peak if possible. Then add the hybrid hardware and only then tune print quality.
cm
cm3w ago
So I took your advice and tuned tensions and added the hybrid stuff. The resonance graphs are here. If I slow eveything down a lot I can get decent results. But if my perimeters are anything over 100mm/s ish the ringing is awful. I also attached my printer.cfg. I set the max acceleration in the ratos.cfg to be 6400mms/^2. Is it possible that the config is just not using input shaping?
cm
cm3w ago
Here’s a benchy I just did
No description
No description
No description
cm
cm3w ago
To test, I ran a ringing print "with" input shaper and with the input shaper commented out. The prints look basically the same. What am i missing? @miklschmidt is it possible that my input shaper isn't being used and its not working?
No description
miklschmidt
miklschmidt3w ago
Your Y belts are overtightened and possibly not equal. Open the realtime analysis tool and create a macro called "Hybrid Y Belt Tension" with the following sequences Sequence 1 gcode: MAYBE_HOME recording off Sequence 2 name: Right Y Belt gcode:
G1 X550 Y250
GENERATE_RESONANCES AXIS=Y
G1 X550 Y250
GENERATE_RESONANCES AXIS=Y
Recording on, use toolhead adxl Sequence 3 name: Left Y Belt gcode:
G1 X-50 Y250
GENERATE_RESONANCES AXIS=Y
G1 X-50 Y250
GENERATE_RESONANCES AXIS=Y
Recording on, use toolhead adxl Save that macro and run it from the resonance analysis tool front page.
cm
cm3w ago
I made it and tried to run it but I got these errors.
No description
miklschmidt
miklschmidt3w ago
Is the configurator up to date? And did you pick the right accelerometer to record with?
cm
cm3w ago
Yeah its up to date and i picked T0 which is right. It gives me the same error when I got Tools-> home printer Same thing when I try to see the live accelerometer streaming
miklschmidt
miklschmidt3w ago
Please post you debug-zip (link in sidebar) Wait, do you have beacon configured as your accelerometer on t0 ? That won't work (yet). Use the EBB42 (select it in the hardware configurator, you don't need to save it to your config)
cm
cm3w ago
No im pretty sure im using the EBB42
miklschmidt
miklschmidt3w ago
yep, confirmed This is very odd klipper is reporting adxl errors
cm
cm3w ago
Hmm. Any suggestions?
miklschmidt
miklschmidt3w ago
no unless you're familiar with web development, i need to see the requests it's making it's definitely requesting the wrong adxl and i just tested the setup here, works fine. Can you double check the settings in the hardware configuration?
cm
cm3w ago
Yeah let me double check those settings
miklschmidt
miklschmidt3w ago
Btw you're getting temperature warnings on your X stepper driver Also show me those macro sequences because it looks like the first sequence completes just fine
cm
cm3w ago
No description
No description
cm
cm3w ago
All setting looked right btw
miklschmidt
miklschmidt3w ago
Yeah this looks correct. But then what the heck is this:
ebhooks client 4133774840: Disconnected
ADXL345 finished 'toolboard_t0' measurements
Generating oscillations at (470.000, 30.000, 20.002)
toolhead: max_velocity: 800.000000
max_accel: 8535.000000
minimum_cruise_ratio: 0.000000
square_corner_velocity: 5.000000
Disabled [input_shaper] for resonance generation
Testing frequency 114 Hz
toolhead: max_velocity: 800.000000
max_accel: 9100.000000
minimum_cruise_ratio: 0.500000
square_corner_velocity: 5.000000
Re-enabled [input_shaper]
Completed resonance generation for axis=1.000,1.000 axis
webhooks client 4128781624: New connection
BatchBulkHelper start callback error
Traceback (most recent call last):
File "/home/pi/klipper/klippy/extras/bulk_sensor.py", line 45, in _start
self.start_cb()
File "/home/pi/klipper/klippy/extras/adxl345.py", line 268, in _start_measurements
raise self.printer.command_error(
gcode.CommandError: Invalid adxl345 id (got ff vs e5).
This is generally indicative of connection problems
(e.g. faulty wiring) or a faulty adxl345 chip.
webhooks client 4128781624: Disconnected
ebhooks client 4133774840: Disconnected
ADXL345 finished 'toolboard_t0' measurements
Generating oscillations at (470.000, 30.000, 20.002)
toolhead: max_velocity: 800.000000
max_accel: 8535.000000
minimum_cruise_ratio: 0.000000
square_corner_velocity: 5.000000
Disabled [input_shaper] for resonance generation
Testing frequency 114 Hz
toolhead: max_velocity: 800.000000
max_accel: 9100.000000
minimum_cruise_ratio: 0.500000
square_corner_velocity: 5.000000
Re-enabled [input_shaper]
Completed resonance generation for axis=1.000,1.000 axis
webhooks client 4128781624: New connection
BatchBulkHelper start callback error
Traceback (most recent call last):
File "/home/pi/klipper/klippy/extras/bulk_sensor.py", line 45, in _start
self.start_cb()
File "/home/pi/klipper/klippy/extras/adxl345.py", line 268, in _start_measurements
raise self.printer.command_error(
gcode.CommandError: Invalid adxl345 id (got ff vs e5).
This is generally indicative of connection problems
(e.g. faulty wiring) or a faulty adxl345 chip.
webhooks client 4128781624: Disconnected
Ah, you're testing the oscillator here it seems
cm
cm3w ago
Yeah I think I was trying different things to see if anything worked
miklschmidt
miklschmidt3w ago
I can't find anything related to that failure. Very odd. Will have to look more into this tomorrow, sorry! in any case, it's prolly worth checking you driver cooling, it's complaining quite a lot it's also reporting openload in some cases, suggesting a bad connection to the stepper. This is not resonance btw, that's artifacts from a stepper acting wonky same with the majority of this Could actually all be down to that one driver struggling from a poor stepper cable connection
cm
cm3w ago
Hmm ok. I have everything stock and the fans are running. Maybe I need more fans? I can check that out Is there a way to diagnose which stepper might have the bad connection? No problem, thanks for your help!
miklschmidt
miklschmidt3w ago
i think it's just the cable having a loose connection
cm
cm3w ago
Does it say which stepper? I assume x because thats the one you said that got hot?
cm
cm3w ago
Ahhh now I see what you were saying
No description
cm
cm3w ago
So after watching the log files, it seems that i get this overheating error randomly for other steppers too. I checked and know my conenctions are stable. Everything is securely in there. This is making me think there is a firmware issue maybe?
No description
miklschmidt
miklschmidt3w ago
This is making me think there is a firmware issue maybe?
No, this is reported directly from the drivers. The "OpenLoad" is a loose connection in the motor cable. Check the crimps Did you forget the driver duct in the electronics ompartment? And are you absolutely sure the fans are running (and running the right voltage)?
cm
cm3w ago
Ahhh the driver fan duct was not on I can check all of the crimps. In fact let me redo them to organize the wires. I will let you know if this fixes it
cm
cm3w ago
What does is mean when it says TMC 'Stepper" reports GSTAT?
No description
cm
cm3w ago
@miklschmidt Undid the belts, re-aligned everything. Also checked all connections. Here are the tension graphs before and after adding the hybrid. Still can't get the realtime analysis to work. Same error as before.
No description
No description
miklschmidt
miklschmidt3w ago
Internal driver error, see if a power cycle helps. i'm sorry i haven't had the time to dig further into it, i still can't reproduce the problem on my end
cm
cm3w ago
No worries, thanks for the help. I think this last print I did was a lot better. Now just battling VFA's
cm
cm3w ago
So looks like I got it all fixed. For anyone in the same boat, I took off the hybrid belts and made their tensions perfect. Then added hybrid belts and tightened them super slowly. I only looked at the shaper graphs instead of the tension graphs. I guess I misunderstood that the tension graphs aren’t helpful for hybrid.
No description
TheTik
TheTik3w ago
gorgeous result!
miklschmidt
miklschmidt3w ago
Excellent sleuthing, this looks great 👍
Dank God Precious
Very nice I'm kinda in the same boat
cm
cm2w ago
Did you get yours sorted out?
Dank God Precious
Nope not yet it's my first diy kit so it's a hella of a learning curve
cm
cm2w ago
Well best of luck, I can try to help if you have questions!