I’d like to start a general discussion about how to optimize a printer regarding to eliminate sounds & noises.
I think almost everybody has already thought about this topic and some of us developed solutions and strategies to avoid sounds & noises.
My thought is to collect them in a central spot as a reference work, where you can find a solution or strategies you haven’t thought of, but might apply well to your setup.
The idea came with the closing words in another Topic
Well that triggered me to reduce the overall acoustic emission of my printer, but where to start…?
So i would like to ask for your insights and experiences:
And if you carefully read the linked articles, both options are not really recommended if you aim for maximum precision and quality.
Whereas stealth chop is considered far worse compared to interpolate.
I think we often use a “bad” combination of TMC stepper drivers and motors.
For example if I fill out the Excel spreadsheet of my TMC5160 to hit the rsense value I have to power the motors with more than 2 amps though their peak is only rated with 1.7 amps.
But with this current the motors are awful loud! When reducing the current they become more silent.
Those TMC drivers are more appropriate for bigger motors or ones for higher speed requiring higher currents and maybe I would be better with TMC2209…
Due to motor resonances the noise often depends on the speed they are running with.
Just a thought on that: Maybe it is possible to measure motor resonances with the accellerometer and avoid the speeds causing the worst resonances in the slicer or maybe in the printer config.
Having a tool like the resonance measurement for the whole printer, but for the stepper resonances also might come in handy.
Of course this is your individual decision but I’d not like to see this as a general recommendation.
Fully agree. The current designs of the TMC5160 are actually pretty crappy for 3D printing purposes as hardly any printer use a RMS current of 3A.
This leads to a reduced ability of the driver to control the current. I had a call with the Analog Devices representative of our company and he confirmed that this is not ideal but that the drivers have internal measures to compensate for such “marketing driven bad design decisions”.
Allows to do some very basic optimizations of internal registers in the drivers. According to my experience:
has no noticeable effect
gives everybody the cozy feeling of having optimized something that hardly anybody really understands
For sure, i think the whole topic is an individual decision to find a balance.
As mentioned in the start of the topic: This is meant as a collection of things which possibly help to find that balance - not as a list of fundamental recommendations.
I put some cheap, Chinese, integrated servomotors (IHSV 42-40-07-24) into my printer as a test and found them to be absolutely silent. At normal print speeds, all you hear running is the hot-end fan and the sliding for the corexy mechanism on the linear bearings.
However, using servomotors will tempt you to use high speeds, jerk, and acceleration, but all three work against silent operation. At high speeds (~200mm/sec and up) the belts make zipping noises as they ride on the pulleys. That can be reduced by twisting the belts so that smooth sides ride on most pulleys, but you can’t eliminate the noise from the drive pulleys. Also at high speed/jerk/acceleration the mechanism will operate quietly on long straight runs but will bang every time direction changes abruptly.
The particular motors I used were not readily tunable for optimal performance and yielded relatively low resolution prints. Better motors that come with proper tuning software might be able to solve that problem. I tried 5:1 belt reduction to improve resolution, and it improved as expected, but still didn’t match the 400 step/rev steppers originally used in the printer, so I put them back.
I use the same servomotors in my corexy sand table that runs up to 1500 mm/sec. With larger drive pulleys I could probably run the thing at 3000 mm/sec. At 1500 mm/sec the mechanism (not the motors) is pretty noisy, but under 500 mm/sec is quiet enough for use in my livingroom. I originally built the mechanism using cables instead of belts but that requires very complex motor mounts with lots of additional pulleys and the ones I used (little V groove ball bearing pulleys) made a LOT of noise (weird clattering noises as the pulleys turned). It was also a nightmare trying to string the cable everywhere and maintenance/modification was very difficult.
Yes, there is a turning point somewhere below 200mm/sec where things become noisy. I’ve limited top speed to 300mm/sec and also observed
So one thing you can do about abrupt direction changes is: Designing your parts with rounds instead of edges (if possible) to avoid the problem. And rounds don’t mean polygons with less than 180 edges on a 20mm diameter.
B.t.w.: Sometimes parts look quite nice(er) if (re)designed with this objective in mind!
Just a dumb question: What is that? How is it used and what for?
That’s something I’ve already heard from ballscrews, when the balls are clattering through the nut. I think that could be solved with less viscosity grease.
I try to use fillets in my designs as much as possible, but sometimes you just can’t do that.
The sand table is my coffee table. There is a 590 x 980 mm corexy mechanism moving a magnet under the table, and baking soda on top of the table with a steel ball that traces patterns in the baking soda. It is lit from two sides at a low angle using LED strips. The LEDs light up the ridges in the sand created by the ball, and cast shadows resulting in a very nice patterns and textures to look at, both while it is drawing and when it finishes. An additional benefit is that high speed erase patterns keep my cat entertained. Patterns are created using a program called Sandify. I do some post processing of the pattern files to allow one speed for the drawing and a much higher speed for motion along the edges of the table.
Example pattern that takes about an hour (IRIC) to draw:
I have blog pages on the servo motor test in my 3D printer and on the sand table design, but this site rejects links to google blogs, so you’ll just have to search for it. The latest version of the sand table is called “arrakis 2.0” (what else would you call it?). If you want to see it drawing I also have a youtube channel with some videos of the table running.
