Input shaper values - FW 6.5.3
I just updated to the final version of 6.5.3 (from the RC release).
Default values for input shaper are X - MZV 60 hz and Y - MZV 48 hz.
I ran calibration with the accelerometer and get values of X - MZV 38 hz and Y - MZV 41 hz, which I have saved.
The X value specifically is quite a variance to the default value. Is this normal to see such large differences to the default position?
RE: Input shaper values - FW 6.5.3
At this point, it would also be a good idea to post your issue on GitHub ( https://github.com/prusa3d/Prusa-Firmware-Buddy/issues). You could do so here in the forum as well, but on GitHub, Prusa can forward it as an internal ticket if your information turns out to be important.
Mods for Core One: Core One HT 450 degrees, Comfortable display , Very fast print start and Reducing noises
Mods for Prusa XL: Very fast print start
RE: Input shaper values - FW 6.5.3
It would help if you could confirm that belts are tuned with the strobe tool, maybe also that the gantry is squared.
RE: Input shaper values - FW 6.5.3
Also, is this a factory assembled printer or a kit? If a kit then there is no way the input shaper defaults are accurate.
RE: Input shaper values - FW 6.5.3
Thanks for the responses all.
My Core One was a kit (Mk4S conversion) and I have subsequently carried out the "+" upgrade.
Belts were tuned via the firmware tool and gantry is square.
RE: Input shaper values - FW 6.5.3
I get 46 & 46 MZV on my Core One kit. There's a lot of cabinet vibration that goes on during this test. Is a conversion mechanically the same as a Core One in terms of housing and everything else? How solid is your bench? Does the accelerometer attach firmly without rattle? I can't believe it makes much difference, but in the spirit of "test it the way you use it" I calibrate with the door closed and filament loaded.
[It seems like X and Y should be different, but those are the numbers I get very consistently.]
RE: Input shaper values - FW 6.5.3
Unless you see obvious ghosting and ringing I wouldn't focus on the numbers too much.
RE: Input shaper values - FW 6.5.3
Bench is fairly solid and accelerometer attaches firmly. I do have the Core One sitting on a placemat so I can move it easily if I need to get to the back connection. When I just had it directly on the bench, the rubber feel stuck hard and it was extremely hard to move. Maybe the setup has something to do with it, but it still seems to be printing very well.
RE: Input shaper values - FW 6.5.3
I have mine on folded sheets of paper towel for the same reason. There's no way I could slide it directly on the rubber feet. I wonder what's actually being measured as it seems implausible that the X and Y resonances would be the same.
RE: Input shaper values - FW 6.5.3
I wonder what's actually being measured as it seems implausible that the X and Y resonances would be the same.
I have been wondering too. The default values, with sixtysome Hz in X, are (qualitatively) what I would expect. But I don't think I have read from anyone who actually got that value when running the measurement.
Maybe it is still buggy, somehow messing up the A+B and A-B motor movements?
RE: Input shaper values - FW 6.5.3
Resonant frequencies come from mass and spring stiffness. I was hoping to find that there's some fundamental difference between the stiffness of X and Y that would compensate for the difference in mass and make the frequencies about equal. I keep looking at the belts constraining each axis but I'm not smart enough to see it. Can anybody else apply more intellectual horsepower? It needs to be stiffer in the Y direction to make sense, since that's the direction with more moving mass.
RE: Input shaper values - FW 6.5.3
In the X direction, only the weight of the print head moves, in the Y direction, the weight of the portal on which the print head is mounted needs to be added to the weight of the print head. That is why the differences in frequencies arise. What is there to not understand?
RE: Input shaper values - FW 6.5.3
In the X direction, only the weight of the print head moves, in the Y direction, the weight of the portal on which the print head is mounted needs to be added to the weight of the print head. That is why the differences in frequencies arise. What is there to not understand?
Yes, that's what we all would expect to find. But in most Core Ones units, the input shaper calibration typically results in no significant difference in frequencies. Both X and Y come out around 45 Hz. Please re-read the posts above.
RE: Input shaper values - FW 6.5.3
same here. Prusa core one + kit, very poor print. input shaper set frequency X MZV 35Hz, Y ZVD 41Hz, with this value the print is decent/good, but the X axis show little wave.
RE: Input shaper values - FW 6.5.3
It would be great if you could give your conditions as well, bench type/stiffness, confirm belt tuning. That X number is surprisingly low, but everything is consistent with numbers well below the defaults.
I'm tempted to put a heavy towel and a brick on top, just to see how much the cabinet is affecting the numbers, or if they mostly react to the stiffness of the core system. One more project for the future.
RISPONDI: Input shaper values - FW 6.5.3
The printer is placed on the floor. Belt tension reads 98/92, but that’s using the Prusa app — I can’t really get along with the built‑in system on the printer. It passes all the tests.
RE: Input shaper values - FW 6.5.3
In the X direction, only the weight of the print head moves, in the Y direction, the weight of the portal on which the print head is mounted needs to be added to the weight of the print head. That is why the differences in frequencies arise. What is there to not understand?
Yes, that's what we all would expect to find. But in most Core Ones units, the input shaper calibration typically results in no significant difference in frequencies. Both X and Y come out around 45 Hz. Please re-read the posts above.
My kit version had very similar values to the the std numbers aswell, didnt make any visible difference to change them.
RE:
I only notice ghosting and ringing on really glossy high temp filaments like PC Blend Jet Black. This is even after recalibrating the input shaper. I haven't seen any difference in 6.4.0 and 6.5.3 even though X and Y (or A+B and A-B) were swapped around
RISPONDI: Input shaper values - FW 6.5.3
In this photo you can clearly see how the X‑axis lines are completely irregular. This happens with the original input‑shaper values and the Speed profile. With the new values the situation improves, but the issue is still present.