Ringing Part 7: The story so far

Things I have definitely fixed

Rods are locked down.
Rods are parallel and not bent.
Belts are as tight as they can be without warping the frame.
Belts are tensioned in a rigid way, and not with springs or other methods that introduce backlash.
Z tower is braced.
Loosened the bearing blocks to reduce friction along Y axis caused by misaligned bearings.
Replaced the Y belt idler with a better-supported version. The old one was twisting a little, and might have been affecting the smoothness of the belt.

Things I have improved, but not fixed

Moved the Y bearings around to reduce friction on the rods.
Lubricated all rods and lead screws with dry PTFE.

Vibration analysis

I’m now focusing on what I think is the source of the ringing, which is the vibration of my build platform caused by the bearings and maybe the stepper motor of the Y axis. I bought a vibration analysis app from the App Store and have begun using it to measure the vibration on the bed by laying the phone face-up on it and performing some full-length moves along X and Y at 3000 accel/10 jerk. Here’s how it looks currently:

I think this is the ringing I’ve observed. As the X axis moves, you can see that the bed is vibrating; it’s going back and forth along Y, and up and down along Z (remember that the phone is face-up on the build platform). The fact that Y is moving back and forth would explain why I still get some ringing on X movements, although not as much as I get with Y or XY movements.

Ah the Y axis. It’s making the bed rattle quite violently. The actual vibration is difficult to see with the eye, but can be felt with the hand. It shakes along X, which explains the more visible ringing, and it also shakes up and down more violently.

Vibration causes

Like I said, the bearings and the stepper motor.

I have printed PLA bushings with large tolerances, and reamed them out more fully with an M8 all-thread rod. I’ve also sprayed then full of dry PTFE, and I’ll install them and re-measure the Y axis motion vibrations. Hypothesis: The not-quite-smooth bearings are creating jerkiness and uneven friction on the rod, which creates vibrations. Therefore if I replace those bearings with smooth-running bushings that have very little friction, then bed vibration should be reduced. And on the same note, if the bushings create extra friction then they should create more ringing.
The motor may be jerky, and it may be caused by the lower baud rate of the i3 (115200 baud versus the PBSM’s 230400 baud) creating some lag between stepper motor instructions, and so the printer stops for a millisecond to wait for instructions. Therefore if I print from SD card, I should get less vibration than printing from USB.

Other possible fixes

People sell after-market Y carriage plates to deal with the fact that the spring tension from improper bed levelling can actually pull the carriage plate up at the corners. These plates are sometimes heavier, and a heavier plate would dampen vibrations quite nicely.

Maybe I can do an experiment where I just add weights to the one I have now, see if it’s worth anything? Good excuse for buying some fishing weights from Kmart…