Prusa-Firmware-Buddy
Prusa-Firmware-Buddy copied to clipboard
prusa3d
[BUG] Y-axis calibration fails
Please, before you create a new bug report, please make sure you searched in open and closed issues and couldn't find anything that matches.
MK4
5.1.0
Original
USB drive
The Y-axis calibration fails. The status line shows y=211mm. I downgraded to 5.0.0, performed the full Self check -> everything passed.
Exactly the same here. I had already taken the beast out of the enclosure, checked all the bearings, tensioned the belts - and then it dawned on me that this is not a problem of my printer, but another serious problem of the 5.1 firmware.
Had the same issue with belt tension setting at 95. When I adjust the belt tension setting to near 98-99 printer pass the y-axis test. Belt need to be very tight.
Well, I got desperate and downgraded to 5.0.0. Self-test passed on 5.0.0, no issues with the Y axis. I will not touch 5.1 again...
Same here. I tired also resetting to factory defaults like suggested on reddit. Tried multiple times. Y axis moves to the very front bumps into end. - Calibration failed. Going back to the 5.1.0. alpha. Acutally it behaves the same - leaving the bed after bumping into the limit covering the display - BUT the test passes.
I have same issue with Y axis moving to very front and failing calibration. Downgrading to 5.0.1.
Didn't have any calibration issues with v5.1. Sounds like what happened with the upgrade to v5.0 though, people with improperly assembled printers getting new errors with firmware that's more strict.
@tkircher so I guess you're saying that Prusa improperly assembled my MK4?
For me, after upgrading to 5.1 GA, calibration randomly succeeds on the Y axis (meaning it randomly fails too). Same symptoms as @pjcar2000 where the plate moves to the front (after successfully calibrating the start) and immediately fails (without the normal retry).
(my printer was factory calibrated, working fine on 5.1.0. Alpha)
I tightened the Y belt all the way to the limit in the tuner app. No change. Then I removed the front bracket, holding the guide left rod which is the end stop, when the LMU8 hits it. Also no change. From this, I figured out the "movement range" might be too large on my printer. In the manual they tell you to assemble the LMU8 slight aligned to the inside. Which results in probably 1mm more travel. I moved the "front" LMU8 to the center of the bracket. This did the trick. Manual page here: (https://help.prusa3d.com/guide/7-y-carriage-heatbed-assembly_438310#438972) Would be super nice to see WHY the self-test fails in the first place. Guess I could have looked in the source then.
BEFORE
FIXED
The interesting bit here is the fact that calibration fails on machines that comply exactly with Prusa's specifications - mine, for example, is factory-assembled, not built from a kit. This definitely can't be categorized as "user errror" or "bad assembly". It's a bug in the calibration routine, plain and simple.
same here.... Y-axis calibration fails and then keeps bashing the frame for while.... happening before prints randomly for no reason... never had any issue before 5.1.0.
Another one here! I never had a Y-axis issue on my MK4 kit on previous firmware! Although it prints very good on 5.1.0 it was aware of this self test issue. Downgrading again to 5.0.1 seems to solve this self test issue. I don’t disagree with @tkircher about fine assembly but if Prusa wants to make the firmware more strict or we should do as @dr-welt proposes I would like to be told so from them!
Thank you!
Another one here! I never had a Y-axis issue on my MK4 kit on previous firmware! Although it prints very good on 5.1.0 it was aware of this self test issue. Downgrading again to 5.0.1 seems to solve this self test issue. I don’t disagree with @tkircher about fine assembly but if Prusa wants to make the firmware more strict or we should do as @dr-welt proposes I would like to be told so from them!
Thank you!
Agreed, Prusa needs to mention it at the very least. And they definitely didn't.
Well, at the very least I'd Like prusa research to acknowledge the issue and say when they are going to fix it. It seems to me that 2 weeks is long enough time for them to make such a fix if they adopted it as a priority. Meanwhile they seem content to leave users in limbo, having advertised new features of 5.1 FW, but folks unable to take advantage of it, because they had to downgrade. Really this is not good enough.
Please, before you create a new bug report, please make sure you searched in open and closed issues and couldn't find anything that matches."
Yeah sure. I did that and nothing came up. The search algorithm on this site sucks OK. Don't blame the users, blame the original developers that made a site for reporting bugs that cant find previous reports. Instead talk about creating a mechanism that allows similar bug reports to be consolidated into 1. Or bring it into the current day and age by adding an AI.
We have to look into some enhancements of the other printers Y and Z test. It could interfere with the axes test of the MK4. bryn51 I have to agree with you in all points. In the past they tried to sort the issues per printer type and issue type, not talking about the release version, you you could a bit easier to fulfill their original issue creation request : First look at the existing issues ... To do so, they really should make the search a bit more user friendly. Labeling it they would even have better overview for themselves. To change the status from open to work-in.progress or setting the expected resolution time is somewhere in the dreamland.
My suggestion is, do not look at any HW adjustments, the mass appearance while in the 5.0.1 everything works ok indicates that it is a real FW bug of the 5.1 version
I tightened the Y belt all the way to the limit in the tuner app. No change. Then I removed the front bracket, holding the guide left rod which is the end stop, when the LMU8 hits it. Also no change. From this, I figured out the "movement range" might be too large on my printer. In the manual they tell you to assemble the LMU8 slight aligned to the inside. Which results in probably 1mm more travel. I moved the "front" LMU8 to the center of the bracket. This did the trick. Manual page here: (https://help.prusa3d.com/guide/7-y-carriage-heatbed-assembly_438310#438972) Would be super nice to see WHY the self-test fails in the first place. Guess I could have looked in the source then.
I hit the same issue and went back and left a comment on the build instructions for the MK3 -> MK4 upgrade. At least it’ll be a heads up for whoever hits it next.
I can verify this bug on my MK4? Please escalate a fix. Belt tension and y Axis bearings are all within build recommendations. Increases or decreases to belt tension did not resolve the self test failure.
Please, before you create a new bug report, please make sure you searched in open and closed issues and couldn't find anything that matches.
MK4
5.1.0
Original
USB drive
The Y-axis calibration fails. The status line shows y=211mm. I downgraded to 5.0.0, performed the full Self check -> everything passed.
Version 5.1.2 of the Mk4 firmware is recently released. I have installed the new version and I have verified that this reported issue remains. I am now contacting Prusa Support to request an explanation as to why this issue has not attained a high priority to be fixed, since it seems likely its experienced by many Mk4 owners.
OK, so 5.1.2 version of firmware arrived. I installed and tested it. The reported issue with Y axis failing self test remained.
I got onto Prusa Support (Andres). While calming listening to me harangue about why this issue had not been fixed, he calmly suggested I look at alignment of the Y axis bearings, namely the two on the left. The assembly guide says they need to be slightly offset towards the middle, but do not say how much, except to say the bearing must not overlap the depression. The assembly guide provides no measurements that would help assemblers get it right. You have to judge by eye in reduced light whether the bearing overlaps the edge of a black hole beneath it. I measured mine using a feeler gauge, and found there was no overlap, but the gap was very small ~0.1 mm. But after the chat closed, I went about the process of re-aligning those two bearings. I did it to test Andre's assertion that an adjustment to bearing position will cure the problem. This task not for the faint of heart or those short on time, as this involves at minimum removal of the heated bed shiny rails, the y axis idler, & the y axis motor before you can easily get at the screws holding the bearings. I nudged one of the bearings a tiny bit, so its enough to pass a 0.2 mm feeler gauge. Then reassembled the whole thing, and adjusted belt tensions during the process because no doubt the previous setting was disturbed by all this (it was). I also tuned X axis belt tension at the same time (It was also maladjusted).
And, I found that the 5.1.2 firmware now passes all the geometry self tests, but I cannot for the life of me understand why it checks the Z axis 3 times over. I'll not be convinced that this firmware change (in the 5.1.0 release) was well thought out or executed or tested. And why it has not improved in the 5.1.2 release.
The 5.1.x firmwares clearly have a much tighter tolerance for the Y axis geometry than earlier versions. But One wonders at the sanity of Prusa releasing firmware that requires every Mk4 Kit assembled owner to rejig their Y axis for the sake of a 0.1 mm tweak escapes me. And all without telling anyone. they ought at least own up to this debacle and apologise to all the customers inconvenienced this way. I personally hold the view that the firmware ought to be re-engineered to have a slightly wider tolerance, not requiring owners to pull their printer apart for the sake of some nuance in exactitude that isn't explained or described in release notes.
Indeed. If we were talking about tweaking a screw or something, one could accept that easily. Being forced to perform this kind of disassembly on an otherwise flawless, factory assembled unit that is just a few weeks old... No, sorry, that's not OK.
After changing the xBuddy board from version 2021 to 2023 (yes, it's strange, it was sent to me by Prusa support) because of an incorrect temperature measurement. After installing the board, the Y self-test failed. After contacting support, I was suggested to downgrade to 5.0.1 and wait until a solution came from the research department. With the downgrade to 5.0.1 the Y self-test worked. I was curious that there was a clue with the orientation of the bearings. The first attempt to increase the distance was by pushing the bearings together as far as possible, which didn't fix the error. Then I pushed the bearings apart to shorten the Y path by about 1mm and it worked reproducibly for me with firmware 5.1.0 and 5.1.2. I didn't change the belt tension. could you check which xBuddy board the error occurs with?
SRY I write German with Google Translate
@blue0ceanby Ich kann Dir die Version nicht wirklich sagen - kann man die rausfinden, ohne das Gerät zu zerlegen? Die Seriennummer ist jedenfalls "10589-3742441631731617".
(I can't tell you the version - can that be determined without disassembling the printer? The serial number is "10589-3742441631731617".)
you have to loosen the 4 screws from the xBuddy board cover then it can be read in the center of the board
Yeah, I'd have to pull it from the enclosure for that... Sorry, that would be a bit too much right now.
No problem, and I would have loved to help you - but the way Prusa designed their enclosure makes getting the printer out of it quite an adventure. :D
So instead of internally escalate, they have a problem, and ask the programmer to sort out what issue they have, they tend to let the users to re-adjust even the factory assembled printers. What happens if the user fails? Will the warranty remain still valid? I think they approach will change only when the factory assembled printer users will start mass sending their printers back to factory to re-adjust on Prusa cost.
After changing the xBuddy board from version 2021 to 2023 (yes, it's strange, it was sent to me by Prusa support) because of an incorrect temperature measurement. After installing the board, the Y self-test failed. After contacting support, I was suggested to downgrade to 5.0.1 and wait until a solution came from the research department. With the downgrade to 5.0.1 the Y self-test worked. I was curious that there was a clue with the orientation of the bearings. The first attempt to increase the distance was by pushing the bearings together as far as possible, which didn't fix the error. Then I pushed the bearings apart to shorten the Y path by about 1mm and it worked reproducibly for me with firmware 5.1.0 and 5.1.2. I didn't change the belt tension. could you check which xBuddy board the error occurs with?
SRY I write German with Google Translate
The x-buddy version does not matter. The FW version is the key. You could earn this issue having one of the 5.1.x FW flashed on your new x-buddy. As far as I will have a time I will look at the firmware, what they did.