How to interpret this resonance graph?

Basic Information:

Printer Model: Kobra 2
MCU / Printerboard: Trigorilla_pro_b_v1.0.2
Host / SBC: Rpi 3 B
klippy.log: not needed…

Graph in Question:

My accelerometer is mounted with all axis’ aligned correctly, no doubt about that. For reference, this is my X axis chart

I’ve run this multiple times on both axis’, this is the most recent result.

I’ve been told to take the recommended input shaper and value, but also to only mind the Y axis and just put the peak in my cfg file… so Which should I do?

*clarification- I’m trying to figure out what input shaper value to put ion my .cfg file, currently it states:

[input_shaper]
shaper_freq_x: 78
shaper_type_x: ei
shaper_freq_y: 64
shaper_type_y: ei

The graph of the Y axis is confusing me because of the z axis resonance during bed movement, not sure whether to use the recommended input shaper+value or to just input the peak y axis value like I currently am running (italicized in .cfg above)

Who told you so?

As this is a bed slinger and in general for other kinematics you want to measure and set resonances and input shaper for both axes, X and Y!
It is true, that the axis with the least suggested acceleration in general dictates or at least suggests the maximum acceleration for printing though you still measure both axes.
If the graphs would be clean with single sharp peaks you want to try the input shaper with the least or in best case no remaining vibrations and smoothing and set those values according to the documentation for both axes:
https://www.klipper3d.org/Config_Reference.html?h=input#input_shaper
That means you define the input shaper type via shaper_type_x and shaper_type_y and set the respective shaper frequencies then restart your printer.
It is often desirable to use MZV if suggested but sometimes EI will be better.

For your Y graph. You wrote that the ADXL is aligned with the respective printer axes. That would mean you have some serious resonance in Z when the bed is moving.
Additionally both graphs are a bit low in terms of PSD. Maybe you want to retension the belts.
This often cancels noise and produces cleaner graphs and improves the effectiveness of the input shaper algorithms.

Check the rods and bearings of the bed and clean those.
Check for smooth movement of the bed and for any loose parts and wobble you want to fix.
Then measure again and post the results here if you want.

1 Like

Maybe Interpreting the Input Shaper Graphs will help to get a general understanding.

2 Likes

Sorry for the flawed writing :sweat_smile: I made it sound like I was going to only put the Y axis input shaper into my cfg file, so I added clarification into the bottom of the post.

I see what you mean with the PSD magnitude, I’ll try retention of the belts first and update if anything else happens.

I believe one of the dips in my Y axis graph shows binding like stated in the Identifying Mechanical Problems section so I’ll definitely check that.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.