RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Completely agree, it is all just subjective thought and conjecture but that’s how all good hypothesis start and are proven or disprove. Unfortunately in a scenario like this you are dealing with too many variables for it to be easily solved by the community. You’re right, that what might solve it for one person may not for another. Prusa is the main hope that they identify the variables through experiment design and systemically rule in/rule out the causes, and come up with a solution… And then they’ll sell the solution to us in the form of a Core One S upgrade kit 😂
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Prusa wants me to send mine to PrintedSolid for investigation (I’m in the US). I’m waiting on the details to get it to a shipper, but they seem to want hands on units exhibiting this behavior.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
In my mind, that’s tantamount to an admission that there is indeed a hardware issue.
Could potentially open the floodgates for similar noise complaints?
Prusa wants me to send mine to PrintedSolid for investigation (I’m in the US). I’m waiting on the details to get it to a shipper, but they seem to want hands on units exhibiting this behavior.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
In my mind, that’s tantamount to an admission that there is indeed a hardware issue.
Could potentially open the floodgates for similar noise complaints?
Prusa wants me to send mine to PrintedSolid for investigation (I’m in the US). I’m waiting on the details to get it to a shipper, but they seem to want hands on units exhibiting this behavior.
Maybe. That or they’re sick of talking to me on chat / unable to identify additional stuff for me to check remotely.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
The "funny" thing is, my core one didn't have any vfas or weird stuff till after a few prints. It suddenly started after a while. I tightened all the screws and checked on the belt tension, but I still got vfas afterwards. Even the inputshaper with the accelerometer and relubing everything didn't help that much.
But then I checked the printed parts on the printer itself and they all had these artefacts too. So maybe I was just very lucky in the beginning.
Noisewise everything seems alright with my printer.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
The "funny" thing is, my core one didn't have any vfas or weird stuff till after a few prints. It suddenly started after a while. I tightened all the screws and checked on the belt tension, but I still got vfas afterwards. Even the inputshaper with the accelerometer and relubing everything didn't help that much.
But then I checked the printed parts on the printer itself and they all had these artefacts too. So maybe I was just very lucky in the beginning.
Noisewise everything seems alright with my printer.
Did you print on structural profile?
It mostly happen when printing straight walls parallel to X and Y axis at 80mm/s (inner walls on structural profile).
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
It changed in the middle of a print, and yes, it's mostly in the X and Y axis. I'm mostly wondering why it changed while I was printing something. Like the first 10mm of the part in which it changed looks perfect, and suddenly then there are artifacts.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I was talking about resonance noises.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I was talking about resonance noises.
Cwould be you were lucky in the speeds and directions of movement on the first parts you printed?
Im hopefully that it is something in the firmware for the x motor control logic, like thenmotor needs to change the number of microsteps or mode for that speed. The noise sounds a lot like a stepper motor that is getting angry about something it is being made to do.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I was talking about resonance noises.
Cwould be you were lucky in the speeds and directions of movement on the first parts you printed?
Im hopefully that it is something in the firmware for the x motor control logic, like thenmotor needs to change the number of microsteps or mode for that speed. The noise sounds a lot like a stepper motor that is getting angry about something it is being made to do.
If it’s really related to motor, guys who said that changing the bad one solved the noise for him could be right.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
just got my Kit last week and i have no problems with noise...
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
Even when printing rectangle shape at structural profile speed?
just got my Kit last week and i have no problems with noise...
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
It seems pretty clear from multiple reports that a resonance occurs at specific print head speeds (40 and 80 mm/s). Has anyone measured the acoustic frequency of the resonant noise? Is it 20 and 40 Hz respectively, which would point towards the belt teeth? Or is it a much higher frequency, which might match the motor steps? 400 steps/rev, 32 mm pulley circumference (?) should translate into 500 full steps/s at 40 mm/s, so 500 Hz or 1 kHz frequency.
Anyone with a well-working musical instrument tuner or with a musical ear who can listen and compare the pitch?
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
There's audio spectral analysis tools, eg spectroid for android phones, that could be used to compare spikes/resonance from different users. Ideally the microphones used should also be identical (same brand same model) but for our purposes it probably doesn't matter and ppl could just use whatever smartphone they have.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I think knowing just the fundamental frequency would go a long way in identifying the source.
My gut feel is still that there may be unit-to-unit variations between different printers on two levels: (a) how strongly vibrations get excited, e.g. due to belt or stepper motor tolerances, and (b) how strongly some components resonate along.
RE:
Anyone with a well-working musical instrument tuner or with a musical ear who can listen and compare the pitch?
Better, I play guitar through my PC with plugins etc. and also used to do some basic mixing/production in Cubase, so I've got some really nice EQ plugin: FabFilter!
I've recorded this absolutely wonderful video (to make it less boring, I didn't fix the focus) and played it back through FabFilter, which has a cool "freeze" mode, where all frequency maximums get frozen, to give an image of the full track.
I'm pretty confident to say, that it's 246 Hz:
I'll probably need a few days to get my mini bassshakers ready to playback a sinewave with the resonance frequency to hold them against all parts of the Core One, trying to find the ones starting to "sing along".
What might be interesting: I just checked my belt tunings and they dropped from 82/86 iirc to just 77/80.
Gonna re-tune them and run the same G-code (attached):
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
I'm pretty confident to say, that it's 246 Hz
Thank you very much -- nice data!
Assuming that was at 40 mm/s travel speed, it's pretty exactly half of my estimate for motor-step-induced excitation frequencies. While I am at a loss about the potential origin of the factor of 2, the frequency is clearly too high for the belt teeth to be the cause.
This might be good news. If the vibrations correspond to the motors' full steps, that means that smoothing out the full steps by micro-stepping does not work well, i.e. the microsteps do not produce evenly spaced motor movements. Which is exactly the issue that phase stepping calibration should address once it is implemented robustly, right?
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
If it's exactly half it could be a subharmonic which can be excited by the higher frequency you estimated (one octave higher -> factor of 2) - just speculation on my part.
RE: Prusa Core One not usable inhouse... Sorry, but the sound was to intensiv
My video and G-Code is alternating between 40 mm/s and 80 mm/s (2 walls, perimeters set to 80 mm/s, external perimeters set to 40 mm/s).
I agree, it might be that 246 Hz is the resonance frequency of "something" of the printer and a signal one octave higher still excites it.
If 40 mm/s means factor of 2, it would explain, why the sound at 80 mm/s isn't as clean and has some "shimmering" in it, due to being factor of 4.
It also makes sense, that some users report very positive results, when replacing one motor.
I sadly can't easily reach the electronics of my Core One, so unplugging the motors, while running my custom G-code tends to be difficult.
For what it's worth, I just squared my gantry as best as I could (again) and tuned the belts. They are now both at exactly 85 Hz and some movements of my G-code sound a bit better, but some cause some "buzzing". Probably because the Y-bearings run "cleaner" now, but can vibrate within their tolerances...
I'm gonna run my G-code again and play a little bit with the belt tensioning. Maybe uneven tuning or higher/lower tension makes a difference.
RE:
I'm gonna run my G-code again and play a little bit with the belt tensioning. Maybe uneven tuning or higher/lower tension makes a difference.
Buzzing goes away at even the slightest uneven tuning. I moved the print head a bit to the front, until the belts hit some resonance frequency and were ringing pretty cleanly. Higher notes are a lot easier to tune by ear, so I got my belts a lot closer to being identically tuned, than at full length.
However I now had issues with endless banging and the only solution was to make the right belt a little bit tighter. Belt tuner still says 85 Hz for both, but at the shorter tuning position, I can hear the right belt being a tiny bit higher pitched.
No banging, no buzzing.
I also played around with higher, lower and uneven tuning, while running my G-code and there's no difference at all regarding resonances. Which makes sense, if one or both motors are causing the vibrations and it's simply the natural frequency of some parts or combination of parts of the printer.
I think we're close to "solving" the puzzle. If we get rid of the vibrations that cause the nasty resonances at 40/80 mm/s, we might improve the VFA a lot too.
Different/better binned motors and pulleys might solve both issues.
I really have to unplug one motor at a time, while running my G-code...