After firmware and slicer updates, first layer won't stick?
I did the updates a day or two ago and printed some small parts ok. Now today I trued to print a larger part and the printer probes 35 places as opposed to 12 on the smaller parts. And now the first layer won't stick and I had to cancel the print 3 times in a row. I thought I read somewhere on here yesterday about similar issues but I cant find it now. I think some people went back to firmware version 5.0.1 to fix it. What is going on? This kind of stuff gets really annoying as it worked on small parts yesterday and all I did was try a larger part that I successfully printed weeks ago under earlier versions of firmware and slicer revisions. And the filament is the same too and I am using PETG.
Thanks
RE: After firmware and slicer updates, first layer won't stick?
Ok, so I downgraded the firmware back to v5.0.1 and I have the same problem So that wasn't it. I also ran the full calibration and fan and heaters test and all passed. What I did notice is that the sliced print when i t starts to print does some weird kind of zig-zag motion as opposed to straight lines? It was the first time I have seen that on a first layer. It also seemed kind of fast for a first layer. I am at a loss for what to try next. Help. 😋
RE: After firmware and slicer updates, first layer won't stick?
I guess I will take out the black PETG and put the orange PETG back in so when I check for a partially blocked nozzle I can see the burnt stuff if its the cause. I will report back later. . .
RE: After firmware and slicer updates, first layer won't stick?
Just a followup. I am waiting on some cleaning filament that will arrive later today so I can clean my nozzle as I quit using my MK4 until it gets here.
That also gave me some time to update my Creality printers so I have a usable/reliable printer when my most expensive one isn't working. And that seems like it is going to be the normal experience from what I am I starting to realize. ☹️
RE:
Well son of a b*tch . . . I figured out my problem for the most part but it brings new questions to be asked. I ran the cleaning filament through the printer and it didn't really appear to have found any debris so I loaded back the black PETG (Overture brand) that I was using and I printed a totally different file first as a test and it came out fine. So I tried my print that was giving me all the trouble and it was the same, giving me all the same trouble. I had printed this print successfully in the past although in orange PETG but that didn't seem to be the problem. So I looked at my stored files on Prusalink and I noticed I had 2 of the file in question so I loaded up the older one and it is printing fine right now although the first layer had a little glitch but I let it run and watched it and it will just have a little defect on the bottom. So my question is can the PrusaSlicer create faulty sliced files? Is it possible that my old file was sliced in an older version of PrusaSlicer and the newer version messed up? How can I compare the files to see the differences? The one thing with the bad file I noticed but couldn't understand was that it started printing in a different section as the one that works, and on the first line that is a diagonal line, it was printing like tiny little circles to make the line as opposed to a straight extrusion like a regular line usually prints. Has anybody run into a slicer outputting a defective g-code before?
RE:
There was a version (or two?) of mk4 firmware that randomly corrupted gcode going to PrusaLink. All kinds of odd things can happen, depending on exactly what lines get corrupted. So, that's a possibility if the older file was transmitted using one of those firmwares. Here's a thread about that, with a link to some others:
https://forum.prusa3d.com/forum/postid/669070/
As long as both your old and new files are text gcode (not the new binary .bgcode), you can use a text file comparison utility compare the two gcode files side-by-side. For Windows, the open source WinMerge utility (winmerge.org) is pretty good. I like one called "Beyond Compare", but it is a commercial product (not expensive, but not free either).
RE: After firmware and slicer updates, first layer won't stick?
Thanks scottw, I feel better just knowing that. In my case the g-code that worked was made in September and the new one made in November is the one that was the problem. Of course with one or two firmware updates and one or two slicer updates I guess it will be hard to figure out exactly what happened. I will go read the info from the link you posted. I appreciate it. I feel better knowing my printer is not the issue. 😀
There was a version (or two?) of mk4 firmware that randomly corrupted gcode going to PrusaLink. All kinds of odd things can happen, depending on exactly what lines get corrupted. So, that's a possibility if the older file was transmitted using one of those firmwares. Here's a thread about that, with a link to some others:
https://forum.prusa3d.com/forum/postid/669070/
As long as both your old and new files are text gcode (not the new binary .bgcode), you can use a text file comparison utility compare the two gcode files side-by-side. For Windows, the open source WinMerge utility (winmerge.org) is pretty good. I like one called "Beyond Compare", but it is a commercial product (not expensive, but not free either).
RE: After firmware and slicer updates, first layer won't stick?
(@scottw)
Did you buy the standard or pro version of beyond compare?
Thanks
RE: After firmware and slicer updates, first layer won't stick?
I just used WinMerge and it says both files are identical but I pulled them from my pc. So I guess I need to pull them from the usb stick in the printer to see if they got corrupted during the PrusaLink transfer. That is likely what happened. I will report back later when I get this done.
RE: After firmware and slicer updates, first layer won't stick?
(@scottw)
Did you buy the standard or pro version of beyond compare?
Thanks
I'm using Standard, which has done everything I ever needed. The free WinMerge is pretty good, so you might give it a try first. When I compared them (a few years ago), Beyond Compare was a bit faster, more "polished", and handled some of the comparisons better. But I have not compared the two products recently.
RE: After firmware and slicer updates, first layer won't stick?
I just used WinMerge and it says both files are identical but I pulled them from my pc. So I guess I need to pull them from the usb stick in the printer to see if they got corrupted during the PrusaLink transfer. That is likely what happened. I will report back later when I get this done.
Sounds like the transfer corruption issue. I had not really followed the issue for a while (I tested PrusaLink early on, but then switched to Octoprint -- so have not used PrusaLink for a long time). Reading the GitHub, it sounds like people are still reporting problems, even with recent firmware. Seems like Prusa needs to add a checksum of some type, so the printer can validate the received file (and ask for re-transmit on mismatch).
RE:
Ok, my report shows *only*14,792 differences found and the only one difference I noticed right away was supports was turned on with the one that was failing although I normally never use supports on this particular print. It will take a while to make sense of all the differences as I am not a g-code expert. . . yet. 😋 Do you have any search g-codes that will narrow it down?
Thanks
RE: After firmware and slicer updates, first layer won't stick?
so user error
RE: After firmware and slicer updates, first layer won't stick?
Why would you say it is user error? Both of the files from my pc were exactly the same when I checked them in WinMerge. But when I pulled the two files from the usb stick after they were transferred via Prusalink they were different. Obviously Prusalink or firmware corruption during the transfer.
so user error
RE: After firmware and slicer updates, first layer won't stick?
Do you have any search g-codes that will narrow it down?
Not really, but have you tried looking at them in the Gcode Viewer that comes with PrusaSlicer? You might be able to look at each file, layer by layer, and see if you can spot something.
RE: After firmware and slicer updates, first layer won't stick?
Yes, that is what it looks like. At least my printer is ok. 😊
I just used WinMerge and it says both files are identical but I pulled them from my pc. So I guess I need to pull them from the usb stick in the printer to see if they got corrupted during the PrusaLink transfer. That is likely what happened. I will report back later when I get this done.
Sounds like the transfer corruption issue. I had not really followed the issue for a while (I tested PrusaLink early on, but then switched to Octoprint -- so have not used PrusaLink for a long time). Reading the GitHub, it sounds like people are still reporting problems, even with recent firmware. Seems like Prusa needs to add a checksum of some type, so the printer can validate the received file (and ask for re-transmit on mismatch).
RE: After firmware and slicer updates, first layer won't stick?
Not yet but I will check that out soon.
Thanks
Do you have any search g-codes that will narrow it down?
Not really, but have you tried looking at them in the Gcode Viewer that comes with PrusaSlicer? You might be able to look at each file, layer by layer, and see if you can spot something.
RE:
Since I had not used PrusaLink for a while, I decided to do a test... PrusaSlicer 2.7.0, MK4 firmware 5.1.0, MK4 on wired Ethernet.
I sliced a pretty complex model, creating a 112MB Gcode file. I had PrusaSlicer save it locally, and then send it directly to PrusaLink-MK4 (which has the USB stick plugged in that came with the MK4). Pulled out the USB stick and brought it to the PC for comparison to the locally saved file. Files were identical ("binary same" from comarison).
One test isn't very meaningful, of course, other than to say the problem doesn't occur all the time. But there are enough recent reports on the Github issue to make me believe the problem has not been 100% solved.
RE: After firmware and slicer updates, first layer won't stick?
I used the g-code viewer and the facts speak for themselves. Here is the first layer with the crazy pattern that was added and failing the first layer compared to the way it always used to print.
RE:
Yeah, I guess I was one of the unlucky ones and it didn't happen to but one file so I guess I will have to just watch for it. If I have a problem in the future I will pull the transferred files from the USB stick and check them that way for the true results. At least now I know what happened. The files were identical on the pc but not on the transferred files. And I never added supports and they were turned on by the transfer with Prusalink. I guess they were the major part of the fail but with all the changes it would have just gotten worse. Maybe I way lucky in that sense. 😎 Mine were regular g-code and not bg-code.
RE:
Since I had not used PrusaLink for a while, I decided to do a test... PrusaSlicer 2.7.0, MK4 firmware 5.1.0, MK4 on wired Ethernet.
I sliced a pretty complex model, creating a 112MB Gcode file. I had PrusaSlicer save it locally, and then send it directly to PrusaLink-MK4 (which has the USB stick plugged in that came with the MK4). Pulled out the USB stick and brought it to the PC for comparison to the locally saved file. Files were identical ("binary same" from comarison).
One test isn't very meaningful, of course, other than to say the problem doesn't occur all the time. But there are enough recent reports on the Github issue to make me believe the problem has not been 100% solved.