I also made request for mesh fade. This would address the warp issue & is available in prusa slicer, but not BS afaik. Not sure why.
Mesh fade ? there is already an extrapolation (and interpolation), I think this is what you are talking about ? problem is that there is no probing on the borders, so it rely on extrapolation wich is not very precise
What im talking about is a feature that basically puts a raft under the part that gradually smooths out the warp until it is flat & then the part is printed on top.
1 Like
ha ok, yes thatās a pretty cool feature that I have looked for in bambustudio, it is normaly (well at least with marlin) handled by the firmware with the GCODE M420 S1 Z10 (for a fade on 10mm ), depending on what they implement in their firmware it make works to add this in the start gcode from within the studio
At the end of :
EDIT : for more informations see Bed Leveling State | Marlin Firmware
1 Like
Haha, I just realise that your idea is great, you print a support material or PETG plate of 5mm with a ABL fade of 5mm and then you can print above this a perfectly flat PLA object ! , never thinked of that before
The ABL does not measure the edges, but extrapolate the mesh data from the inner parts. But if the inner parts are halfway flat and the edges are not, it cannot work.
1 Like
Yes, yes, I posted an issue on github about this some weeks ago, they will change that soon : Bed levelling should make measurements on the borders of the plate (and may be more measurements) Ā· Issue #1559 Ā· bambulab/BambuStudio Ā· GitHub
For those with significant machining and metal working experience, Iād like your thoughts. I was looking at the edge of the aluminum plate from my original bed.
It sure looks like it was stamped out, rather than routed out.
Probably with some sort of edge finishing after stamping.
The website compresses the picture too much for me to be able to tell. Iāll probably take my bed apart in a week or two. May be water jet or laser cut too. I wouldnāt expect them to mill out the plates shape honestly. No reason to be that precise.
This would be great to have in Orca.
2 Likes
Yes, I plan to make my own magnetic glass bed. But also my own fiber plate bed.
I was going to try to make the fiberplate magnetic on both sides so I can avoid using any clamps. Worth a shot.
Well, you can already do a try in both Bambu & Orca slicer using the M420 gcode
1 Like
have you actually checked the measurments results?
each point mapped at room and at high temperature?
the changes are not the biggest and again it is the edges that is the problem due to the plastic frame not beeing aligned.
yea when heated things warp, but please look and compare how much! sure, if you are looking for Micrometer accuracy , this is true, the bend is bending, but we dont see any milimeters hereā¦
We are actually seeing changes in the range of up to 0.3mm when heating, if the measurements above are precise, more on the edges then the center.
And 0.3mm is not a huge dealbreaker compared to nearly a full millimeter, but people had complained about values much smaller then that already.
I managed to get my bed from 0.7-0.8mm warp down to 0.08mm just by uninstalling it from the printer. It must have had quite some mounting stress to flatten itself out so much and the new bed I installed did not have such problems. I still believe its more a QC problem then a general design problem, else every single bed would be affected, which clearly is not the case.
Itās 100% stamped. Very obvious if you look on the corners.
much valid point there!
As you say, it is mainly the initial bending, i do like your comment about pretension in the support frame.
I am still manually waiting for a new magnetic surface after manually making the metal sheet straight and shims out the plastic edge that on my both bed was a huge issue (0.5-1mm)
but i will try remounting my frame at elevated temperatures and see if any change
1 Like
Tested GCODE M420 S1 Z10, and it seems to not works.
I have sended an issue, as it would be a great feature.
Another hint that the firmware is not just a stolen Marlin 
2 Likes
Multiple people have confirmed, with measurements, that the plate changes at temperature.
Right, the changes arenāt the biggest, but they are still changes.
Since the plastic frame has a higher coefficient of thermal expansion than the metal heat plate itās screwed into, itās expanding ever so slightly more, which is causing minor variations in the surface.
This is about what Iād expect when doing the math. Iām not sure what exact plastic the frame is made out of, since I canāt find any stamp marks (maybe someone can check their removed bed and find it for me), but Iād assume itās a plastic that has a CTE of about 75 (aluminum is about 20).
256mm x (75 x 0.000001) x (60C - 20C) = .76mm total linear expansion
The smaller aluminum plate is only expanding about .23mm at the same temperature (again just an estimate since I donāt have exact measurements of it). The plastic frame is pulling at the thin aluminum plate and causing thermal stress and variations in the surface that are fractions of a mm.
2 Likes
Aaaaaaaaaand with these last few comments Iām out of my league and will let the learning process begin
You guys are sure smart and Iām enjoying reading all of the test results while waiting for a viable solution to the issue
Thanks everybody for your knowledge, testing and input
With all of the talent chiming in weāre bound to find a good solution
1 Like