RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Is there some G-Code, yet, that moves the print head in a pattern that produces the noises, but won't make it hit the stops, if you unplug the motor?
Probably two files, one moving the print head back and forth in X and one in Y, in the middle of the printing area. Unplugging one motor makes the head safely move diagonally for both files.
40 mm/s is by far the worst for my Core One. 80 mm/s second worst. Strangely, 120 mm/s and 160 mm/s are quieter, than 60-70 mm/s. Apparently, whatever is causing the noise, only really resonates at lower frequencies.
Just to throw in my "magic speed" again for others: 136 mm/s in stealth mode is by far the quietest speed and for me, a good mix between efficiency, layer adhesion and dimensional accuracy. The belt ripple "VFAs" are also pretty good with it.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I really want to buy one, but if it makes the same engine noises that sing like a CR-10 from 5 years ago, no 21xx drivers and no exchangeable EZ2209 driver, no 64-bit CPU, no EMC module exchangeable from 8-16- or 32-GB EMMC.Too bad for 1300€ not to have EZ2209 drivers and 64-bit CPU. Maybe 1350€ then and the nozzle stuck at 290° not 320° original with nozzles that cost 35€
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
and no noctua fan kit as an original option there are caps and t-shirts on the pursa site.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I worked through removing, testing, and looking at a number of bits with support today. Pulleys, belts, etc., but nothing found. They want me to take out the belts and steppers to look at the pulleys. I was looking through the assembly guide and I'm not sure I'm up to it. A huge amount of the printer has to come apart.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I haven't really measured, but in my experience the main thing that makes it louder than the mk4 (before I converted to Core One) is the chamber fans, that can indeed be loud when it's trying to cool down the chamber.
Besides that... There is the fact that it's faster, so faster motors will indeed make more noise. Structural profile (instead of speed) is less noisy.
Then again, the printer is in my garage, and the sound doesn't bother me at all.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
One of the drivers for me picking the core one was the fact they specifically call out the low noise in the advertisement. Still waiting on my Kit and quieltly hopeful people are just overly used to how quiet the MK4 supposedly is. I don't think with a Core XY it would have matched it but not great so many are calling this out. Has anyone put a decibel meter to it, phone app or dedicated device, to record what it's peaking at from a few feet away? I know in the advertisements they have it peaking out at about 50dB, are people seeing well above this?. I currently have an ankermake M5 in my office and that thing is obnoxiously loud, has a sustained 60-70dB and could peak up to 75 on some fast movements. When I put a paver and foam under it, it drastically isolated the higher pitch vibrations from the faster movements which meant I could hear it less around the house. However with the fan whine still means I can't comfortably have video meetings while it's running, which is generally how I spend half my work day, so frustrating that I don't run it as much as I'd like to. Was really hoping Core One met its sound promises so I could actually print more frequently. From what I hear still a lot quieter than most of the other enclosed pre-build Core XY options
RE:
Here's a video of my Core One in Stealth Mode and XY limited to 136 mm/s. The infill prints at 136 mm/s too (0.25 mm nozzle). Chamber fans are limited to 38%, which doesn't make much difference to the default 40% regarding airflow, but already reduces the noise quite a bit.
I installed the app "Noise Meter" and my Samsung S20 FE (400€ in 2021) shows 45-58 dB, depending on the angle and distance to the printer.
It's definitely louder, than the print head of my mk3s ever was, but also definitely quiter, than the bed movements of my mk3s, when going beyond 60 mm/s.
40 mm/s and 80 mm/s have nasty resonances, but anything else below 136 mm/s isn't louder than the video. And since I've run the phase stepping calibration, from 136 mm/s up to 400 mm/s, just the "whooot" and Y-axis bearings noises become gradually louder.
I'll try to do another test with the 0.4 HF nozzle in the next days and post a video + noise meter levels too.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I'll try to do another test with the 0.4 HF nozzle in the next days and post a video + noise meter levels too.
Could you try to print an object with flat vertical walls and a big round wall please?
I couldn't get my printer to avoid those resonance noises (at 40 and 80mm/s) because I wanted to print slow enough to have strengh, and even if I printed outer walls at 100mm/s, a big round wall made the print head travel at 80mm/s (straight axis speaking) at some point of the round edge which lead to the annoying noise.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Sure I can. Do you want to upload a zip with the 3mf, so I can record the noises for direct comparison and then also just use the stl with my own speed settings?
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I will do that yes! 🙂
Thanks!
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Sure I can. Do you want to upload a zip with the 3mf, so I can record the noises for direct comparison and then also just use the stl with my own speed settings?
Here is the link of the STL and the Gcode of the structural settings printed object to test. LINK
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
You didn't adjust the Start G-code to disable the heating during homing/nozzle-cleaning(/probing) and I just watched my print bed trying to find to nozzle in thin air for 2 minutes with multiple retries and banging the print head in the corner, to then fail, lol.
Auto Home without heating worked flawlessly though... As expected.
So I losened the nozzle thumbscrews a little bit, gave the hotend a little wiggle and tightened the thumbscrews again, carefully -> Works
(multiple attempts for all Z-axis things though and without the custom G-code workaround, I would "send this piece of expensive garbage back").
Anyway, beautifully printing now!
I also tried to keep my phone with the Noise Meter app about 40 cm away from the top front edge, pointing the microphone to that edge.
As you already said, you can't quiet down the big circles, since each axis will always move 2x through the nasty resonance speeds per circle.
But for my ears, it's fine and not long enough to become annoying. Not great, not terrible.
Your G-Code:
1. Normal mode, 100% speed: 55-61 dB according to "Noise Meter"
2. Stealth mode, 100% speed: 53-60, similar dB, but lower pitched noises
3. Normal mode, 200% speed (became way quieter after the Phase Stepping calibration): 57-61 dB, higher pitched, but not higher dB
And here are my custom speeds with my custom stealth mode (3mf attached in a zip): 52-58 dB, clearly quieter to the ear and also a little bit in dB
Sure I can. Do you want to upload a zip with the 3mf, so I can record the noises for direct comparison and then also just use the stl with my own speed settings?
Here is the link of the STL and the Gcode of the structural settings printed object to test. LINK
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Oh crap sorry for the start gcode, I didn’t change anything.
Your first video seems to be the wrong one, it’s not my STL. Haha
Nonetheless, indeed phase stepping seemed to have work a little, at 200% speed there is no resonance noises even during the circle.
Could you reupload the first video with my STL at 100% speed please?
Thanks a lot!
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Thanks for all that testing. I think those dB ratings are pretty good in all accounts. Particularly if the phone was only 40cm away. Real world scenario I hope won’t have someone sitting that close to the machine. Even just another meter further back and that will lower slightly again as the sound gets bounced around the room. I mean peaking at 60dB that close is really good, considered background chat loudness. Based your data I’m guessing the complaints are less related to the core one being too “loud” but more the peak “noises” are not nice to the ear, harsh tones making it more annoying. Might be a harder fix.
Quick question, was the door open when you were recording the decibels as well or just the video?
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Your first video seems to be the wrong one, it’s not my STL. Haha
Oops.. Here is the correct one.. Simply copied the wrong link, sorry:
No worries about the Start G-Code, it's not your fault, that Prusa screwed up and looking at the nozzle in the wrong angle makes it fail everything... 😀
RE:
Thanks for all that testing. I think those dB ratings are pretty good in all accounts. Particularly if the phone was only 40cm away. Real world scenario I hope won’t have someone sitting that close to the machine. Even just another meter further back and that will lower slightly again as the sound gets bounced around the room. I mean peaking at 60dB that close is really good, considered background chat loudness. Based your data I’m guessing the complaints are less related to the core one being too “loud” but more the peak “noises” are not nice to the ear, harsh tones making it more annoying. Might be a harder fix.
Quick question, was the door open when you were recording the decibels as well or just the video?
The door is closed in the videos too, hehe, I recorded them with the phone about 2 cm away from the door. Thanks to no light source in the room, other than the printer's LEDs, there are basically no reflections.
I agree about the noise level alone not being the big issue. Without Phase Stepping, high speed moves have some "bass" in the noise and the high pitched "whoop" is also more annoying. But even with Phase Stepping, 25/40/80 mm/s create some nasty resonances, that sound like your gantry would want to destroy itself.
The speeds in the structural preset are identical to the structural preset of the MK4s (I think, not 100% sure right now and not at home). So Prusa apparently didn't adjust the speeds to the Core One and now it's unusable for many, mainly due to the 80 mm/s for perimeters.
I really hope Prusa was/is just lacking manpower and it's a matter of time to get things right and not a matter of money or lack of skills..
RE:
Oops.. Here is the correct one.. Simply copied the wrong link, sorry:
[...]
I really hope Prusa was/is just lacking manpower and it's a matter of time to get things right and not a matter of money or lack of skills..
Phase stepping calibration isn't official right now, Danopermis on GitHub gave the gcode line to "beta test" it, but wasn't confident/satisfied to release it officially.
He states on GitHub that he almost regretted giving it because most people started to complain on Reddit instead of helping him.
So I guess it's still a work in progress, and we can hope that the final version will improve noises more than the actual "beta".
I think it's not 100% firmware related tho, I bet there is a design flaw somewhere that is still undiscovered. I really hope it will be fixable (by Prusa or user).
"VFA" is a thing, but the resonance noise is what really bothers me the most.
EDIT: Thanks for the videos and tests! 🙂
RE:
Ya, if the movement is hitting a specific frequency to cause that resonance, it’s unlikely software will completely remove the issue. But it could still markedly improve it alright though. To be completely rid of the underline issue I’d imagine some kind of physical dampening/decoupling in specific areas is needed.
Considering the resonance seems to be worst when the extruder moves straight on each axis, and VFAs are far more pronounced on straight walls. I wouldn’t be surprised if that resonance is the cause or at least some what related to the VFA issue people are seeing. I mean if the whole system is vibrating as it moves you’re going to see it in the print. I’m not a 3d printer but if you violently shake me as I was trying to draw a straight line, the line would end out a little shaky even with my best efforts 🤣
RE:
Yep. To me, it also seems like the pattern of the belt ripple "VFAs" are also causing the vibrations, which result in the resonances at specific speeds.
It doesn't seem very related to the belt tension, but the tension does have some impact. It's not the belts starting to "sing".. It seems more like the tension on the pulleys/idlers changes how the belts are "creating" the vibrations.
Maybe we could sacrifice a bit of printing area and dampen the Y-rods and X-rail? Like screwing some rubber dampers between the uprights and the ends of the rods, interrupting the resonances of CoreXY frame.
Or something like these little feather-weights on bow strings, which dampen the ringing and *ploink*, when shooting the arrow.
Or maybe it would be enough to put some of these car doors damper mats on multiple free areas on the uprights, bottom and CoreXY frame.
When building the kit I was like "Oh cool, I'm building a steel-guitar. Gonna make it play some printing blues". And since putting some gel pads on a drum or taping a tissue on it already works wonders, it might help to do something similar with the steel parts of the Core One.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
To solve a problem it is paramount to identify with near certainty what is causing it, otherwise it's like trying to hit a mosquito with a throwing knife. In the woods. In a moonless night. Sure people can try all possible solutions to all possible problems and by pure luck hit the right one but I wouldn't count on it. Also, what works for one person may not work for another - we can't assume there's only one problem, maybe there's different problems with similar symptoms.
I don't claim to know the source of the resonance some people are experiencing, and VFAs are very hard to see in my prints (you have to look at them in a certain angle against the light). There seem to be some people who have pronounced resonances and others who don't. In my mind that points to two possible causes - either there's different hardware (the belt, pulleys, bearings etc) or it's assembled differently (alignment, angles, screw torque etc) - or maybe a mix of both.
To test that we'd need people from both camps to have a defined setup (what the coreone sits on should be nearly identical) and the same tools to measure things, eg a type of microphone in a defined position and a spectral analysis software (and/or an accelerometer in a certain spot). That would eliminate some of the subjective/localized perception (like I claim I don't hear much from my printer but maybe I'm just a deaf old piece of wood 😉 )