I had been having issues with getting my M4 to calibrate since I finally tried to set it up 6 months ago. This is a M4 kickstarter, but when it finally arrived life got crazy and I couldn’t get to it until about 6 months ago. I spent hours trying to even get it to calibrate and it just kept giving me errors. I have asked in the forums some and people mentioned different things, like the magnets being backward and stuff. Sometimes it would calibrate, but by the time it was done I had to shut it off for the night. The next day it would not accept the calibration and tell me I had to do it again then it would just fail to retract the belts and I would spend hour trying to get one or two to retract. The next day it would retract but give me errors on calibration.
I had all but given up but then saw the 4.1 upgrade kickstarted and decided to wait for that. I received the 4.1 upgrade, took it all apart and re-assembled with 4.1. I made absolute sure I had all the parts together.
It all went together fairly well. I like how the belts fit so much better in the belt guard I decided to try it vertically vs horizontally like I had before. I had my frame at 15 degrees. It has firmware 0.88 on it. In my garage, around 75 F. I hung it, verified it has retraction force at 900, and a 9x9 calibration grid.
I retracted all belts, then extended them, connected to my frame and started the calibration. It did its (what I assume) normal process. It took about an 30 min to get its fitness where it was happy then started the calibration grid. I assume around half way thought the grid when it pulled tension the top right belt snapped right in the middle of the extended distance. It was a pretty violent snap and I think the belt whipped back and hit the head. The belt wrapped in the wire for the top right encoder when it snapped back. It then immediately started giving me an error about the top right encoder failed to read. I have plugged it back in several times and I immediately get that same error.
I am assuming it broke the PCB, because the white port the wire plugs into is loose now.
Has anybody ever had this happen? I still have yet to actually cut anything with this thing. I am a pretty technical person but I am starting to think this thing is cursed. I did save the log from the time it broke the belt if that is helpful.
Any ideas before I throw this whole thing in the trash and just write it all off? I don’t want to but I have spent so much money and literally have nothing to show.
@bar Thank you so much! I really want to get this thing to work. I love fiddly things, I have built plenty of 3d printers and laser cutters and work in concert production with moving lights and PA’s and stuff. But for some reason this just has me beat at the moment. I have spent so much time building and taking it apart and yet still can’t make it work.
I feel this! I had lots of trouble, and because I can’t just leave it set up and tinker whenever I have a moment, I only made progress (so-called) when I actually had a project at the theater or the school where I could set up a frame. Plus occasional carelessness and just bad luck…it’s been an adventure!
I finally sorted many things out during my last project and it did just what I needed it to (4.1-ish hardware and firmware v0.88). It. Was. Glorious!
Have you replaced the index file and the maslow.yaml file? I’ve found sometimes they have gotten corrupted somehow. and Since you haven’t got cutting, I would highly suggest you do the smallest calibration grid because it will calibrate so much faster and get you can get cutting faster. I’m not certain what the benefit of 9x9 would be.. perhaps better accuracy? but just go with with small calibration to get you cutting. How are you connecting to Machine? Android, iOS, windows? Does it make you feel it is a stable connection? Are you connecting direct to the Maslow and going to 192.168.0.1? I find calibration is tough on my cell phone, because of the scree timing out on my phone, and the unstable connection issues messes with the calibration. SO I think having a solid connection method is important. Once you figure it out, it definitely works. It has some nuances (bugs) but they will take time to iron out, because there is so much going on to make it all happen. Upload pictures and videos and your logs, and I’m sure we can all help out once we understand the whole picture.
This is super interesting, I think that you found a bug.
From looking at the log it looks to me like the encoder connection was lost before the belt snapped (and probably lead to the belt snapping).
During normal operation (cutting something) if the connection is lost to an encoder the machine will shut itself off to protect the belts. It seems like that doesn’t happen during the calibration process so losing an encoder will result in a belt likely breaking. That is 100% something that I need to fix.