Voron 2.4 (350) Belt Comparison Issues

[frank thistle]

Voron 2.4 350 (LDO rev C from Fabreeko)
Beacon 3D Probe (USB) w/ ADXL345 (using this to do resonance)
EBB36 (CAN) w/ ADXL345
Reaper Toolhead dual 5015 (v24) fans with Phaetus Rapido v1 and 3030 (v24) hotend fan
wwG2 Extruder with dual runout sensors
Sensorless homing
Galvanic Umbilical holder Mod and rear umbilical pathing (Umbilical has both USB and CAN wires)
#18 piano wire for umbilical strain relief
24V All Motors
6mm A/B Belts, 9mm Z belts
Sunlu Black ABS gantry parts and carraige
Pin mod on A/B and X/Y joints (verified that the idlers spin freely)

I am wanting to resolve the bad input shaping that I've been struggling with and I figured that I should start at the source, belts. I am also curious to find out at what point does the peeks get introduce so I am tearing down the toolhead to the bare minimun and slowely add toolhead components while testing between each step.

The printer has over 1800 hours of print time on it before it fell off my workbench and destroyed the frame. I rebuilt it from the ground up with a new frame. I also replaced the stealthbuner, tap, and drag chains setup to the reaper toolhead, wwG2 extruder, Beacon Probe, and the ebb36 CAN board. With this new setup I have over 217 hours of print time. I have verified that all bolts are tightened and that there was nothing loose.

For the tests I home first then run the belt comparison test, record the results and run it again. I do this three times for each step. Unless specified I will only post the best of the three for each step.

• Control Base test
  • For control I first undone my belts tension and then retensioned them ~1/8 turn at a time until I was able to measure ~2.1 with the GT2 Measuring tool.
  • I homed, ran QL and then homed again.
  • I ran the belt test three times (forgot to run input shaper unfortunately but I do have older ones) without any filament loaded.

• Barebones Base test
  • After the control I did the most barebones test I can do. I removed the toolhead from the carriage and ran a separate USB cable to the Beacon probe.
  • This test should great results for the belt comparison (forgot to run input shaper at this step unfortunately).
    FINDINGS:
    • The test came out almost perfect. The belt frequency profiles were dead on with a best of two peaks with 0 delta on frequency and 3.8% on amplitude delta.
    • I believe this test tells me that the belts and the printer are in good shape, and that any issues that I have from here on out must be from the toolhead.

QUESTIONS: What mechanical aspect of the printer can be tied to the two peeks?

• Hotend test
  • I installed the hotend onto the carraige with the hotend fan. I tested this step two full times, one with wires loose and one with wires tiedup (a total of 6 times, 3 loose, and 3 tiedup)
  • I also ran input shaper callibration test at this step as well.

First is with loose wires and second is with wires ziptied

QUESTIONS:
- Why would the hotend add complexities to the graph in this step from the barebones step? Is it really the wires?

[frank thistle]

• Extruder test
  • I installed the wwG2 extruder without the ebb36. I put in a peice of filament to stop the ball barings in the dual sensors.

Running the belt tests now

[cunning tusk]

I would install a galvanic bowden tube holder
https://www.printables.com/model/978123-voron-bowden-ptfe-tube-guide-arm-and-canbus-cable

Also since you are using the pin mod I would review my video https://youtu.be/3yWxMeZkKp4?si=zZBfOemUe18eAXze

[cunning tusk]

This is what the belts and x-axis should look like without a toolhead on it

[cunning tusk]

This is not a good test to run. This is becaues the weight is not correct. and your toolhead is bottom heavy. It will through off the results

@royal wasp you might be interested in this

[cunning tusk]

@frank thistle Put the entire toolhead on the x-axis and rerun belts and input shaper

Since you have taken everything off lets get a new run

in addition, lets use EBB36 for all future reasonce testing

[frank thistle]

Ok, I will put everything back on and rerun using the ebb36

[frank thistle]

I have the tool head back on and did a QGL. Doing belt tests now

I forgot to change from beacon to ebb36, but here are the results if wanted. I am now running the belt and input shaper with the ebb36 right now

Here are the graphs from the ebb36 runs

I also did watch that pin mod video and I verified that both my X/Y toothed idlers can shift up and down a little

[frank thistle]

@cunning tusk I have an LDO input shaper from my v2.4 LDO kit. I can print out a nozzle mount and use that, would that be any better or will it suffer the same as the beacon would?

[cunning tusk]

Use the ebb36 values.
Set input shaper to ei for x and y

Print a voron cube

[gritty veldt]

@cunning tusk sorry to hijack but any thoughts on why the beacon results are so much worse than the ebb36

[cunning tusk]

This was originally made for someone else, but I think the same thing may apply to you

[mighty basalt]

Any update for us on this?

[frank thistle]

Unfortunately not much. I did use the ebb36 for input shaping but I still see ghosting. I’ll do a check with an adxl on the nozzle to see if I can get some progress.

[mighty basalt]

do u have any update for us?

[frank thistle]

Ive been struggling with ringing for months and It’s honestly demoralizing. It seems like I’ve tried everything. It’s been a great toolhead otherwise. I’m thinking of going back to SB as I didn’t have nearly this bad of a time with ringing but I figured it would be best to try troubleshooting from scratch.

I haven’t done any troubleshooting for almost two months so hopefully a fresh mind/attitude will help. What would you recommend be the first steps to take if I want to start from the beginning?

[mighty basalt]

Checking all toolhead screws
Checking all XY Joints (yes i know u have pin mod but still) according to the attached picture
Check if the motor pulleys have the belts in the middle with equal space top/bottom
Use pf make tension tool for belts AB and Z
Rerun all tests with sweeping disabled #1308097646786777108 message

[lusty junco]

If you're interested in other toolheads for lighter weight and better cooling performance take a look at the A4T. It uses almost all the same parts as XOL but with a vastly superior, more or less monobody design.

Reaper is cool and all but it looks like an absolute brick of a toolhead with all the extra weight from the fans and associated hardware

I actually built one like 80% of the way before dwTAS caught me and pointed me to the A4T

anyway, worth checking it out -> https://github.com/Armchair-Heavy-Industries/A4T

[frank thistle]

Ok cool! It’s totally worth a shot. I’ll get started on printing parts out. Thanks again!

[mighty basalt]

how are things going? do u need further help or can we mark this as solved?