Resonance calculation attributes Y resonance to Z in graph
Hello,
I am measuring Input Shaper using the "Generate Shaper Graphs" script.
The results are below.
But from the images, I see the graph for Y shows lime color as if the sensor thinks that vibration is in the Z axis, is this just a visual bug?
I'd also like to know whether the graphs look fine, or should I adjust tension.
V-Core 3 400 with EBB36 1.2 toolboard.
Thank you!
16 Replies
you can set the axis in the printer config, this depends on the orientation of your AQDXL sensor, but its just visual, works jsut fine as it is
Thank you, that's okay, how about frequency analysis, should I tighten the belts a bit more?
these graphs cant tell you if you need to tighten your belts or not
i would start with the belt tension graph isntead of this ones
once the belt tension graph deosnt show any unwanted resonances then you can start with input shaper
There it is
this is how it should lookl like
your lines are telling me that oyur rig suffers from a lot of resonances
they shoul dbe smooth and not zizagish
also, you suffer from binding somewhere, you should have two main peaks
i woul dcheck the belt path, the belt is binding somewhere, in the idlers or motor cages
also, control every single screw on the toolhead and the printer
i see multiple resonabce sources in your graph
the peak 1t 100hz is likely located at the toolhead
check your fan mount screws
the peak at 100hz should not be there, try to get rtid of it
Interesting... all movement tests work very well, as a matter of fact I have printed two benchys and they came out well, but I'll double-check everything.
well, i dont say you cant print with it, you can. but you can print much faster if you fix the resonances
I printed at 250 mm/s with 4.5K acceleration. Of course, I'll do some more tests, just getting started 🙂
Is this on a 500 ?
no, but it should look like this on all v-cores
this is the goal
after you managed this, then you can start with input shaper, before its almost pointless
Is the 4000000 spectral density and 50 /80 hz ranges to be targeted? On my 5th 500... Still wondering... I start by pre tuning with a piano tuning app.
forgot the values on the left side, ignore them
forget the frequencies
jsut go for the visuals
1. make sure the lines are smooth, not like a wallstreet chart. these are vibrations of your printer. you need to fix this before going on
2. after point 1 is done, try to equaly tension your belts, there MUST be 2 peaks, the first one should be a tiny bit shorter then the 2nd one
thats the trick
after this input shaper
forget sound measuring apps, thats nonsense
That was great on Prusas Bear before input shaper...
yep, it works to a certain degree, but with ADXL sensors that cost 1€ the sound method is obsolete
Couldn't agree more. Piano tuner makes the initial belt tensioning much faster, prior to proper klipper tuning. I've built more than 20 klipper printers over the last 2 years and I've spent countless hours on this. I feel homing XY between each adxl sequence kinda levels the tensions on the belts. What do u think?