I had to make a large, flat pulley like part and when I set it up on my drive motor it was way out of being flat and true! I suspected the bed of the printer was not flat and the bed leveling was not work. Long story short I printed a Dial Indicator holder and tested my bed. Its out by 0.019" in the rear right hand corner! I was able to dial a little of that out but no way I am going to be able to change it with the single adjustment knob in the back, alone. Anyone else tried testing their bed? I had a Sovol 03 and you could dial it in pretty well with a dial indicator but it would not hold it for very long.
You may want to move this post to the “X1” forum category as it’s not really MakerWorld related. There should be a pencil icon next to the post title.
Exactly how did you perform this test, long story?
Bambu had a major problem with truly warped beds produced in the first quarter of 2023. Typically they were off by 0.7-1.1mm, and until late May-early June replacement beds were often also warped. My experience with the problem and the final result is here.
If you ask Bambu support for help, they will ask you to manually tram the bed. (Print a wrench, it makes precise adjustment much easier). That will make the bed as parallel as possible to nozzle travel, but of course will not make the bed flat. They will also want logs from a small print made with bed leveling turned on (a tiny cube will do).
If you use your favorite search engine to find the expected precision/accuracy of FDM printing you will find that 0.019 inch (0.48mm) is right at the limits. AFAIK, Bambu has never said what they consider to be an acceptable level of flatness.
Thanks for this information. I did watch a couple YouTube videos and did print the adjustment wrench. I made my own custom dial indicator holder. As a professional machinist I find it incredible that 0.019" would be an acceptable tolerance for a 3D printer that can print with a .2 mm nozzle and down to 0.08 mm layer height. At least you could dial most of the un-levelness out of my SV03. Why on Earth they designed the adjustments to only have three points is beyond me?
I just used the G-code for leveling. I will try running the dial indicator test this weekend.
There was very little adjustment required but I bet my right rear corner is still off by a mile.
The same way I set up any piece of materials in my commercial CNC Vertical Milling Centers. I run a machine shop for a living making custom parts. I use the 3D printer for pattern making and fixturing.
I meant the exact process you are using on the X1. I’m assuming you have mechanical expertise enough to attach and read a dial indicator. Exactly how are you manipulating the X1 to make the measurements and in what sequence?
What is the starting state of the machine? Power cycled? Or you just finished the tramming g-code? Then you attach the dial caliper to then head, then… (exact steps)
People here are trying to help you, they aren’t questioning your qualifications.
I homed the printer. Removed the hot end, installed my dial indicator and zeroed it where the homing was. That was my X,Y 0 Ref. Then went corner by corner moving the print head with the touch screen, recorded the initial values off from X,Y ref. Then proceeded to divide by half the error across the three adjustment knobs until I was close to X,Y Ref at all 4 corners. The rear right corner was the worst off. That is how I did the error check.
Thanks for the details. Did you do anything to set the ABL mode before this process or did you just power-cycle? If you just ran the tramming script it is out of ABL. If you just printed something from studio it is in ABL. Not sure what the state is on power-up.
You bring up a good point about the Auto Bed Leveling feature but if the base is warped no amount of ABL can make the part flat. When you place a piece of raw material in a CNC machine it still has to be faced before you can get a reference for level for the rest of the machining process. This is additive machining but the premise is the same.
I do not know whether the Homing process with all its checks is the same as ABL but I kind of doubt it. Would require some more testing on my part. But thanks for pointing the issue of the ABL out to me.
I wrote that if you finish printing an object ABL is left on, actually I’m not so certain, I just don’t see it being turned back off. So better to explicitly set it one way or the other if you are running manual tests.
This may be of interest to you:
Next time, after you zero your gauge, shut the printer off. Then, you can move the printhead freely. Makes it a lot quicker/less frustrating.