RE: Constantly recalibrating
I upgraded the FW and started messing with the belts and homing became 12 bangs.
I dont understand the instructions and nobody is willing to explain properly here nor provide a video, so until better instructions are provided I will just stick with the previous Fw.
It's been explained over several threads. Start the calibration. When asked to tune the upper belt, rotate the knob until you see maximum deflection of the upper belt - it'll be a smooth back-and-forth several times a second. Click the knob, then do the same for the lower belt. That's it - nothing complicated.
RE:
I upgraded the FW and started messing with the belts and homing became 12 bangs.
I dont understand the instructions and nobody is willing to explain properly here nor provide a video, so until better instructions are provided I will just stick with the previous Fw.
Prusa has no obligation to provide detailed instructions while their firmware is in alpha. Also, it's not that we are unwilling to explain, we've explained a lot and in detail already. If you are still not able to understand, please just wait until the firmware is officially released. Then if you are still having difficulty understanding the process you can at that time contact support.
RE:
I came across this post, because I am also facing this problem. I am also very sure that I did everything which can be done to ensure the gantry is properly aligned and the belts are tuned as they should be. See my separate post.
I believe there is a non-neglectable fraction which faces this problem for a reason which has nothing to do with improper belt tuning or bad gantry alignment. Yes, for the majority of users the homing procedure just works. Otherwise, Prusa would have already solved the issue. Then there is probably a good fraction of users which faces this problem and where belt tension or gantry alignment are simply bad. Tuning the belts and aligning the gantry makes the problem disappear for them. As this solves the problem in many cases, this might lead to the wrong assumption that this is always the solution. However, I believe there is also a fraction of users who have a perfectly assembled Core One and still see the issue.
After I have read all posts in this thread, two other aspects stand out:
- Temperature of the chamber
- Stability and vibrations of the underground
Before I read the posts here, it has never occurred to me that those both aspects might contribute to the problem. However, the longer I think about it the more I believe that heat might be a problem.
In my original post, which I have linked above, I wrote
Each time, the first auto homing (i.e. directly after the belt calibration) worked nicely: only two bangs and homing was finished. I thought, yeah, that was it. But after the first or second print the homing procedure was running forever again.
I could understand that the belts of a newly built Core One need to settle in, because the belts might give in after some time and needed re-calibration. But that shouldn't happen after a single print.
After the initial assembly the selftest and very first homing procedure went smoothly. Everytime after I repeated the belt tuning, the first homing procedure ran smoothly, too. I re-tuned the belts, because I didn't know anything else to do although the tension of the belt seemed to be fine and exact according to the Belt Tuning app. Everytime, I thought, yeah, finally it works. But then the problem re-surfaced after the first or second print.
However, I am mostly printing high temperature material such as PC. Chamber temperature is between 55°-60°C. Maybe this is the problem. After the initial assembly and after each repeated tuning, the chamber had been cold of course. But I never made the connection.
I will pay attention to that from now.
RE:
same problem here , usually bouncing both corner 8 times wich is already too much , and randomly get into endless calibrating , bouncing 50 times , and bed temp go up crazy because of initial 115degC setpoint ? i'v built my CoreOne 2 weeks ago and this problem happend since day one ... i usually reset and go into AutoHome option
would be good to know how much corner bounce needed when normal ? can you all reply with normal corner bounce please ? belt tension should not be as critical .... i agree with @pcr : possible overkill tolerance needed ? after all , my prints are good so those tight calibration are useless
this is a video about my usual 8 times X-Y calibration before print :
For me it's interesting to see that in the firmware for the XL the tolerance is 0.15mm but for the Core One it is 0.075mm. seems like overkill to have it that exact or?
Printing since 2019 , Ex MK3s owner , current Core One owner
RE: Constantly recalibrating
i'v open a support chat ... keep telling me to re-adjust belt tension ... i'v ask for a return procedure since many customer here never find a solution ... and also show them the abnormal initial bed heat setpoint who involve OVER HEAT and delay for the next print after a reset to allow bed to cool down : this is an extensive calibration bumping and bed setpoint video ;
Printing since 2019 , Ex MK3s owner , current Core One owner
RE: Constantly recalibrating
i may have found a walk around :
when doing belt tension adjustement , ounce they both sound OK , you finish by using X+Y belt tension to PERFECTLY align the gantry against front bump , only half turn of any of those belt will misalign gantry by 0.5mm ! so you fine tune bolt final belt tension as a GANTRY alignement ... will see if this method last for me ....
those CoreOne user who end the procedure without checking this might be the ones who have endless bump ?
Printing since 2019 , Ex MK3s owner , current Core One owner
RE: Constantly recalibrating
i may have found a walk around :
once [both belts] sound OK , you finish by using X+Y belt tension to PERFECTLY align the gantry against front bump , only half turn of any of those belt will misalign gantry by 0.5mm !
I might have found a similar workaround.
First of all, in my case it is definitely temperature related. When I perform a belt tension by following the rules to the book, homing works perfectly after a cold start. As soon as the printer gets a little bit heated, homing fails with endless bumps.
However, I run the Vernier Skew Test. After a perfect belt tuning according to the manual, the Vernier Skew Tests reported a negative skew of -0.5mm. Note, that this does not mean that there is gap of -0.5mm between the gantry and the front-left stop position, it only means that there is an -0.5mm derivation across the entire length of one of the diagonals.
I tuned the belts such that the Vernier Skew Test showed a zero. In my case, I had to loosen the upper belt and to tighten the lower belt until the upper belt's frequency was only 94Hz (instead of 98Hz) and the lower belt's frequency was 95Hz (instead of 92Hz). That gave me some concerns first, because it was the opposite of what the manual says (the upper belt should be tighter, not the lower one) and the Prusa Belt Tuning App was quite unhappy with the result.
But it seems to have done the trick. I have been printing without homing issues since a week even when the printer chamber has been hot.
I believe those settings should be fine. Even though the frequencies are not what the Belt Tuning App wants, both frequencies are still in the range between 92 and 98Hz (even though reverted). Hence, the stepper motors should not become overloaded and the belts not too stressed.
those CoreOne user who end the procedure without checking this might be the ones who have endless bump?
That sounds as if those users did something wrong and missed a step. Those additional steps are clearly not part of the official instructions.
Still while it seems as if (and you) might have found a workaround that works for us, I am still wondering what else might be going on here. It simply does not feel right.
- A total deviation of 0.5mm along a 33cm diagonal is so little that this should not have such an impact on the homing procedure. The homing procedure cannot be such sensitive.
- Let's assume tuning the belts to 98Hz and 92Hz would only be the first step to initially set the belts and get a proper starting position, but the real trick was to make the gantry perfectly aligned in a second run which might even mean that one ends up with 94Hz and 95Hz. If that was true, Prusa should be aware of that by now. Still there is silence from Prusa and/or they keep telling us that one only had to tune the belts to 98Hza and 92Hz. On the contrary, they built a new Belt Tuning function into the upcoming firmware version which uses the stepper motors to make the belts oscillate. This gives me the same wrong and non-working results as the current method.
RE: Constantly recalibrating
I've said this earlier in the thread but thought I'd repeat if it helps others. I cured my persistent constant calibrating by putting it on a VERY stable solid surface AND placing a 10ish kg weight on top of the printer. A box full of filament, that also rests against the wall behind the printer.
I've done nothing else as it prints beautifully - and these changes cured my constant calibration instantly.
I might add that I'm annoyed I have to do this - as it shows a manufacturing defect - I suspect in the rigidity of the case. But for now it prints well so I live with it, but I do feel very different about Prusa .... 🙁
RE: Constantly recalibrating
I've said this earlier in the thread but thought I'd repeat if it helps others. I cured my persistent constant calibrating by putting it on a VERY stable solid surface AND placing a 10ish kg weight on top of the printer. A box full of filament, that also rests against the wall behind the printer.
I've done nothing else as it prints beautifully - and these changes cured my constant calibration instantly.
I might add that I'm annoyed I have to do this - as it shows a manufacturing defect - I suspect in the rigidity of the case. But for now it prints well so I live with it, but I do feel very different about Prusa .... 🙁
since my first print i also felt the top of the printer is shaking alot ! things will eventually get loose inthere !
i may have found a walk around :
once [both belts] sound OK , you finish by using X+Y belt tension to PERFECTLY align the gantry against front bump , only half turn of any of those belt will misalign gantry by 0.5mm !
I might have found a similar workaround.
First of all, in my case it is definitely temperature related. When I perform a belt tension by following the rules to the book, homing works perfectly after a cold start. As soon as the printer gets a little bit heated, homing fails with endless bumps.
However, I run the Vernier Skew Test. After a perfect belt tuning according to the manual, the Vernier Skew Tests reported a negative skew of -0.5mm. Note, that this does not mean that there is gap of -0.5mm between the gantry and the front-left stop position, it only means that there is an -0.5mm derivation across the entire length of one of the diagonals.
I tuned the belts such that the Vernier Skew Test showed a zero. In my case, I had to loosen the upper belt and to tighten the lower belt until the upper belt's frequency was only 94Hz (instead of 98Hz) and the lower belt's frequency was 95Hz (instead of 92Hz). That gave me some concerns first, because it was the opposite of what the manual says (the upper belt should be tighter, not the lower one) and the Prusa Belt Tuning App was quite unhappy with the result.
But it seems to have done the trick. I have been printing without homing issues since a week even when the printer chamber has been hot.
I believe those settings should be fine. Even though the frequencies are not what the Belt Tuning App wants, both frequencies are still in the range between 92 and 98Hz (even though reverted). Hence, the stepper motors should not become overloaded and the belts not too stressed.
those CoreOne user who end the procedure without checking this might be the ones who have endless bump?
That sounds as if those users did something wrong and missed a step. Those additional steps are clearly not part of the official instructions.
Still while it seems as if (and you) might have found a workaround that works for us, I am still wondering what else might be going on here. It simply does not feel right.
- A total deviation of 0.5mm along a 33cm diagonal is so little that this should not have such an impact on the homing procedure. The homing procedure cannot be such sensitive.
- Let's assume tuning the belts to 98Hz and 92Hz would only be the first step to initially set the belts and get a proper starting position, but the real trick was to make the gantry perfectly aligned in a second run which might even mean that one ends up with 94Hz and 95Hz. If that was true, Prusa should be aware of that by now. Still there is silence from Prusa and/or they keep telling us that one only had to tune the belts to 98Hza and 92Hz. On the contrary, they built a new Belt Tuning function into the upcoming firmware version which uses the stepper motors to make the belts oscillate. This gives me the same wrong and non-working results as the current method.
X_Y calibration parameter are definitely too tight ! i told this many time in my support chat but they just dont wanna change anything ? ... my prints are ok so why the calibration fail ?
Belts should have a bigger range of adjustment without causing problem ... that sound test method is ridiculous
Printing since 2019 , Ex MK3s owner , current Core One owner
RE: Constantly recalibrating
My endless bumping/XY calibration still present in random condition , even after multiple belts tunning , gantry alignement , pullys alignement...
i'v been working as automation/electronic specialist into a fast packaging plant for 35 years , and 3D printing since 2019 : please consider my knowledge …
i 3D print in my garage , including winter low room temperature ( 15degC) , i dont want to mess with belt tunning and failed prints because of XY calibraion failed.
CoreOne FACT :
--MANY MANY users have this XY calibration/CornerBumping problem
5 pages thread link here
18 pages thread link here
-- the abnormal/extended 5 minutes calibration CornerCollision/bumping « multiple retry» will eventually dammage/misalign the printer top structure + Axis parts . Even the current « normal » amount of bumping are excessive .
--Most user who still encounter failed XY calibration have good prints thus confirm calibration tolerance being too tight : they still perform good prints
--High speed corner collision doesnt help stable calibration readings? Why not go slower?
--Some initial VERY cold chamber (low rooms temperature) doesnt seem to help
--My first print attempt often fail at XY calibraion , after a reset and AutoHome request , it usually pass and work fine
Core One firmware change suggestion :
--Slow down that corner hit/collision test (the one happend at each prints starts)
--do only 1 or 2 pairs of bumping calibration and simply trigger this message if calibration fail :
« XY calibration does not meet specification : abort print or continue anyway ? »
--Change the initial bed 115degC setpoint when chamber pre heat are requested : this cause bed over heat and can lead to dangerous over heat/bed dammage/fire :
Make the bed initial temperature setpoint same as print bed setpoint until the chamber reach its setpoint
Printing since 2019 , Ex MK3s owner , current Core One owner