inputshaping on IDEX?
-
Hi,
I am currently working on my IDEX again and I am getting heavy ringing at around 100mm/sec and input shaping seems to have zero effect. A test benchy has ringing at 100mm/sec at 2.5mm intervals, so 40hz ringing. Setting input shaping to 40hz MZV or ZVD seems to have no effect at all.I am pretty aware that inputshaping is not a panacea for all sources of ringing but I am a bit stumped.
Here is my current config.g. I am running RRF 3.5b2 on a Duet 3 mini 5+ with two MKSTHR running the experimental 2040 toolhead firmware. Other than the massive ringing, the prints are fine.
; RRF 3.x config for CR20 IDEX - dual independent Z ; General preferences M451 ; Enable FFF mode G90 ; Send absolute coordinates... M83 ; ...but relative extruder moves G4 P5000 ; allow 5 seconds for CAN to settle ; Network M550 P"CR-CAST" ; Set machine name -- CR20 with IDEX and Bear style mods M552 S1 ; Enable network M586 P0 S1 ; Enable HTTP M586 P1 S0 ; Disable FTP M586 P2 S0 ; Disable Telnet ; Drives M569 P0 S1 ; Physical drive 0 goes forwards (U) M569 P1 S1 ; Physical drive 1 goes forwards (Y) M569 P2 S0 ; Physical drive 2 goes backwards (X) M569 P3 S0 ; Physical drive 3 goes backwards (Z left) M569 P4 S0 ; Physical drive 4 goes backwards (Z right) M569 P124.0 S0 ; Physical drive 5 goes backwards (E0) M569 P125.0 S0 ; Physical drive 6 goes backwards (E1) M584 X1 Y2 Z3:4 U0 E124.0:125.0 ; X Y Z:Z U E0:E1 M906 X1200 Y1200 Z800 U1200 E600:600 I30 ; Set motor currents (mA) and motor idle factor in per cent M350 X16 Y16 U16 Z16 E16 I1 ; Configure microstepping with interpolation M92 X80.00 Y80.00 U80 Z1600.00 E631:631 ; Set steps per mm M566 X210.00 Y210.00 Z18 E300.00 P1 ; Set maximum instantaneous speed changes (mm/min) M203 X9000.00 Y9000.00 U9000.00 Z600.00 E6000.00 ; Set maximum speeds (mm/min) M201 X7500.00 Y4000.00 U7500.00 Z100.00 E6000.00 ; Set accelerations (mm/s^2) M204 P20000 T20000 ; Set print and travel acceleration maximums M84 S30 ; Set idle timeout ; Axis Limits M208 X-139.8:105 Y-105:110 U-80:140.7 Z0:200 ; axis limits. Move Z uprights back to regain Y. ; Endstops M574 X1 S1 P"124.io0.in" ; X endstop connected to Z probe port on toolboard M574 Y1 S1 P"io4.in" M574 Z1 S1 P"!io5.in+!io6.in" M574 U2 S1 P"125.io0.in" ; U endstop connected to Z probe port on toolboard ; Thermistors M308 S0 P"temp0" Y"thermistor" A"bed" T100000 B3950 M308 S1 P"124.temp0" Y"thermistor" A"left" B4725 C7.06e-8 M308 S2 P"125.temp0" Y"thermistor" A"right" B4725 C7.06e-8 ; Heaters M950 H0 C"out0" T0 ; heater 0 uses the OUT0 pin, sensor 0 M950 H1 C"124.out0" T1 ; heater 1 uses the out0 pin and temp0 on the toolboard #124 M950 H2 C"125.out0" T2 ; heater 2 uses the out0 pin and remp0 on the toolboard #125 M307 H1 B0 S1.0 M307 H2 B0 S1.0 ; temperature limits M143 H0 S115 M143 H1 S290 M143 H2 S290 ; Fans M950 F0 C"124.out2" ; x part cooling fan M950 F1 C"124.out3" ; x heatsink fan M950 F2 C"125.out2" ; u part cooling fan M950 F3 C"125.out3" ; u heatsink fan M106 P0 C"left part fan" H-1 S0 ; part cooling fan is not thermostatically controlled M106 P1 C"left head fan" H1 T45 ; print head fan is thermostatically controlled M106 P2 C"right part fan" H-1 S0 ; part cooling fan is not thermostatically controlled M106 P3 C"right head fan" H2 T45 ; print head fan is thermostatically controlled ; Bed and tools M140 H0 ; bed heater ; first toolhead (left) M563 P0 S"left" D0 H1 X0 F0 ; first tool, X head G10 P0 X0 Y0 Z0 S0 R0 ; set tool 0 offsets and temperatures ; second toolhead (right) M563 P1 S"right" D1 H2 X3 F2 ; second tool, U head G10 P1 X0 Y0 Z0 S0 R0 ; tool offsets ; -- copy mode M563 P2 S"copy mode" D0:1 H1:2 X0:3 F0:2 ; tool for copy mode G10 P2 X110 Y0 U0 S0 R0 ; tool offsets for copy mode M567 P2 E1:1 ; set mix ratio 100% on both extruders for copy mode ; -- mirror mode M563 P3 S"mirror mode" D0:1 H1:2 X0:3 F0:2 ; tool for mirror mode G10 P3 X110 Y0 U0 S0 R0 ; tool offset for mirror mode (flipped X offset) M567 P3 E1:1 ; set mix ratio 100% on both extruders for mirror mode ; Pressure Advance and InputShaping M572 D0 S0.02 M572 D1 S0.02 M593 P"mzv" F43 S0 ; enable display M918 P2 E-4 G4 S1 ; pause M918 P2 E-4 M150 X2 R255 U255 B255 S3 ; set all 3 LEDs to white ; PID tune M501Here are some test prints in chronological order. The first one was an early one that had various problems, after that one where I increased Y current to 1.2A, then another one with the fixed setup. third and fourth were rotated 45 degrees so the bow of the boat pointed to the front right corner of the bed. This reduced ringing on the backside considerably. The last one also had X jerk reduced from 360 to 210.
I can hear pretty harsh vibrations from ... something on the printer when the X toolhead changes direction, with the reduced jerk it sounds much less like the stepper throwing up. I haven't fully located the source of the vibration yet.
Any ideas how I could improve the prints while keeping the 100mm/sec print speed up that is used for these benchies are more than welcome (-:
-
@oliof said in inputshaping on IDEX?:
M204 P20000 T20000 ; Set print and travel acceleration maximums
How did you calculate these acceleration values (or at least pick them)?
Have you tried something a little simpler than a benchy? I like the ringing tower test as it gives you a chance to test X and Y. A little simpler to start there - quick prints, easier to see the results from changing input shaping values, jerk values, speeds, etc.
-
@sebkritikel I tried simpler things and they have similar issues.
also, the M204 is just a maximum value that I set to a (high) limit. I am not using acceleration control from the slicer, so the acceleration values from the line before M204 are being used
M201 X7500.00 Y4000.00 U7500.00 Z100.00 E6000.00 ; Set accelerations (mm/s^2) -
@oliof Ah totally missed that line!
Is the ringing worse in a specific single axis direction (noticeably worse in X vs Y)? Input shaping working better on a single axis?
-
I have had similar motor-vibration issues with RRF 3.5b2 on my IDEX. -> Forum Post and Video
In my case, print speeds between 30mm/s to 100mm/s produced horrible resonance in the motors. Switching the firmware back to stable helped. Unfortunately, I did not have the time and patience to investigate the cause further.
-
@MaxGyver unfortunately, switching back isnt really viable with the 2040 toolboards.
-
@sebkritikel I will print the standard ringing tower in a hot minute, maybe that will shed some light.
<>RatRig V-Minion Fly Super5Pro RRF<> V-Core 3.1 IDEX k*****r <> RatRig V-Minion SKR 2 Marlin<>
-
on my particular machine with the Klipper test print I see the same ringing at my accelerations, as well as reducing X down to 4000mm/sec^2. The frequencies I measure are pretty low (about 20 Hz) ... I assume on the smaller benchy the repeating pattern double up. My best bet is that this is due to the shoddy table. I am not sure that Input Shaping can deal with frequencies this low.
-
@sebkritikel I did reduce accelerations to 3k on both X and Y and I am still getting ringing artefacts at roughly 3mm frequency with and without IS (100mm/sec print time still). With 3.5-beta.3 there are some other issues when using IS that I need to investigate further. Not sure if this is IDEX related, or just my machine exhibiting bad ringing from a source IS can't do anything about.
-
@oliof My IDEX gantry is fairly heavy - a few kilograms, so I typically run my print moves at 1000 mm/sec^2, and my travel moves at 4000 mm/sec^2 (I do outer walls/perimeters from 45mm/s to 120mm/s, and right now travels at 450mm/s - 600mm/s).
I previously calculated 4000 mm/sec^2 as the max safe acceleration value for the X-U gantry (weight/motor torque, etc), and while I did see substantial improvements with input shaping at that acceleration, I still had (IMO) unacceptable acceleration induced ringing. I settled on 1000 mm/sec^2 based on test prints and measured data from the shaping plugin.
20hz does seem really, really low - reminds me when a bearing block mounting screws were a touch too long, so the mounting plate could actually rotate slightly on direction changes. Once I noticed that, I was able to eliminate that frequency from the input shaping charts, and focus (and reduce) the 49hz band (or frequencies around that, I believe I split the distance to hit two frequencies (X and Y)) with EI3.
-
@sebkritikel my individual IDEX heads are pretty light (two EVA toolheads, can run 20k and more acceleration on other machines) and only run in X (its still an i3 style bed slinger), so I think I am good on the weight front.
I guess I need to actually put an accelerometer on mine since the Klipper test print shows almost nothing in terms of ringing. if that gives me weird graphs, I will go hunt for things that move when they shouldn't.
