MK3S Firware V3.8.0 Calibration Software problem
Following two days of trawling through forums and doubting myself, pulling my printer to bits twice trying to resolve a Z axis calibration failure I turned to the firmware. I regressed my firmware back to an earlier version and hey presto my printer works again. Thanks Prusa software development team for regression testing before releasing to the wild. Two days of my life i cant get back.
Bit harsh that. 3.8.0 is working fine on my two machines and I suspect most others too. Perhaps you have an inherent Z issue that the new firmware is more sensitive to.
Prusa devs - thanks for the firmware. Looking forward to what you do next 🙂
Not harsh at all. If your response be true then let the devs provide more intuitive feed back or serial logs that declare the issue their firmware is aborting the operation on. Its like the Microsoft Error "Something Went Wrong". If your writing software / Firmware for a world class printer then act like it.
Many people here were facing axis problems after flashing 3.8.0 firmware, factory reset helped. Something to do with values stored in EEPROM in different format. Try this and see if it helps, you will have to redo the calibration.
Thanks for the guidance, I tried factory and full default with no result. Only resolve seemed to be firmware regression. Hope they sort it out with some formal advice. Such a nice printer, hurts PRUSA reputation when you leave users to their frustrations.
it needs "factory reset with data clear" not just "factory reset"
. 3.8.1.RC2 has more tests to force the reset
If Prusa devs were smart, they'd force the factory reset themselves if and when one is needed, e.g., any time the location of cal constants moved in EEPROM. Since that is almost always if anything in a define (.h) changes, well: it's clear they don't really understand robust coding practices.
And I have suffered so many regression issues with Prusa I no longer pop in the latest firmware because I just don't need the headaches (yes, I used to do a lot of beta testing). I'm still at 3.7.1.
And how many times have they adjusted crash detection levels (the motor currents they sense) and not tell us in the change release notes? Oh wait, Prusa doesn't have release notes. And if you search github, even the changelogs show that a file was changed, but rarely why. Prusa software does not meet the minimal standards for commercial software development. It is mostly hackware; and is more similar to code developed by non-professions programmers world wide for fun rather than business use.
What they are doing is okay for a startup, kinda-sorta; but at some point they need to turn a corner and step up their practices. Sure, the base software is also open souce, like Marlin, but if you read the Marlin code, it is done quite well, good comments, structure, etc. After Prusa got done with it? Not so much. Changes made with not a single comment as to why; interrupt handling that has been butchered; entire working sections of code commented out (that results in odd behavior). And when you hack other working code, it's even worse than hacking your own code because the reputation of the Marlin code set is now up in the air.
I use the Prusa stuff; I don't hate it. But I do understand the quality is lacking in so many ways. 3.8.1 has many new "features" that are more knee jerk "what can we change so we can say we are innovating" that break long lived features that customers have found quite useful. So why break things that work? Why not focus on resolving some of the long standing problems that have been on the books for months or years? Why add to that issue list?
Pretty sure I'll be pummeled ... but that's okay. Without the people standing up for something like quality companies usually ignore it.