And how did you fixed the glass? With clamps like this?
My other printer is a Bibo 2, which has a very thin aluminum heated bed and relies on a glass plate to provide the flat printing surface. I can attest by the bloody stains on the wall from hitting my head over and over that glass on warped thin aluminum tends to also deform after repeated heating and cooling without a flat base under it. I have struggled greatly with first layer adhesion on that printer and was probably the biggest reason I decided to upgrade to the Bambu X1C. Every time I bought a new flat piece of glass, it only lasted a bit until it too would start bending. I have a jig with a dial indicator set up for that printer, and despite the 3 leveling points being dead on, there are dips and raised areas across the glass +/- .3mm so a span of .6mm.
When you talk about glass, did you ever tried mirror ? I have used mirrors for years on my two other printers with great success (3mm thick mirror 300mmx310mm, about 4€), but sincce I discover the Bambulab textured plate, I finally dont like anymore the glass finish on first layer…
Yesterday, I let my new print bed soak a 60C for a couple hours, then cool, then soak at 100C for a couple hours, then cool. It did not change the X axis warp for a PETG print – still around 0.13mm. The X axis warp for PLA does seem to have improved – now around 0.3mm. Haven’t yet checked the Y axis, that’s on today’s to-do list, as is a full bed adhesion test.
The ABL on the X1 is great, but has problems when outside the outer probe points. There isn’t a whole lot that you can do about this, it’s a side effect of the curve fitting. Well, you could move the outer probe points to the edge of the bed. That would improve things a bit. But given the current points, if you have a bed warp, you will have adhesion issues outside the outer probe points. How much and whether it matters depends on the warp characteristics.
I think 0.3mm is perfectly acceptable, dont you ?
I understand people that want thing to be perfect but really it would be good to separate people with a real curved bed and people with little warping that are perfectly normal on a FDM printer, because if Bambulab receive thousand of complain and if half of them are not justified, they will tend to be less reactive for people with real problems.
The good news is that the probes points should be adjusted in a future release.
EDIT: Bambulab should really add a button to make a bed diagnostic (making measurments at different temperature)
I’m suffering from this. Has someone found a way that we can get access to the mesh level values so I can check where is the low spot of my bed and can maybe try to compensate with aluminium tape?
Good idea with that indicator on the z screw. Did you calculate what degrees of rotation equate to in axis movement?
Whatchya doin’ there? Is there a gcode you could share for that sweep? I just tinkered my bed “stright” and would be curious how much it z’s now.
Yes, I wrote this G-Code for several movements:
M17 X1.2 Y1.2 Z0.75
G90
M83
G28 ; home all axis
G29 ; Auto Bed Leveling
G1 X128 Y128 Z10 ; move to center
G1 X35 Y35 F4000 ; Unten links - oben rechts
G1 X220 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y220 F4000
G1 X220 Y35 F4000 ; Unten rechts - oben links
G1 X35 Y220 F4000
G1 X220 Y35 F4000
G1 X35 Y220 F4000
G1 X220 Y35 F4000
G1 X35 Y220 F4000
G1 X220 Y35 F4000
G1 X35 Y220 F4000
G1 X220 Y35 F4000
G1 X35 Y220 F4000
G1 X35 Y35 F4000 ; Unten links - unten rechts
G1 X220 Y35 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X35 Y128 F4000 ; Mitte links - Mitte rechts
G1 X220 Y128 F4000
G1 X35 Y128 F4000
G1 X220 Y128 F4000
G1 X35 Y128 F4000
G1 X220 Y128 F4000
G1 X35 Y128 F4000
G1 X220 Y128 F4000
G1 X35 Y128 F4000
G1 X220 Y128 F4000
G1 X35 Y220 F4000 ; Oben links - Oben rechts
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X220 Y220 F4000
G1 X35 Y35 F4000 ; Quadrat
G1 X220 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X35 Y35 F4000
G1 X220 Y35 F4000
G1 X220 Y220 F4000
G1 X35 Y220 F4000
G1 X128 Y128 F10000
Did the bed stay flat after the bending and after heating up ?
After bending the bed has a minimal bulge upwards. When it warms up, it becomes almost flat.
Did you bent it cold or hot ?
I do not suggest bending while hot. Magnet rubber becomes crumbly so you will likely damage it.
I bent it when it was still about 30-40 degrees Celsius. So far it looks OK.
Nope. I’d expect it to be acceptable on a $250 Ender knock off. On a thousand plus dollar printer whose manufacturer goes out of their way to highlight the precision of their machine? I would expect warp to be less than half the minimum layer height. I would accept warp equal to the minimum layer height. Since Bambu doesn’t actually advertise such specs, I’ll take the 0.08mm extra fine slicer settings as the minimum. That would put my expected values at +/- 0,04mm across the bed.
For those who will point out changes due to heat. That can be engineered out.
My reference point is my four year old Mk3, which, by most measures, is an inferior printer. Except that is actually has a flat bed. Flat to below the measuring accuracy of the printer – <+/-0.02mm. I fully expect my Mk4 to be equal or better.
If anyone has access to a precision lapping stone 12inx12in or larger this would be easy. Just lap the top side of the metal plate back and forth on it til flat. If the plastic housing is distorting the plate (kinda doubt it) you can liquid shim it before tightening screws. Hoping I can find a large enough lapping plate to try somewhere.
Thanks for the breakdown photos and information! Regarding the aluminum PCB, is the 3mm thickness the overall between the FR4 “heater” board + the aluminum? Or is the aluminum itself 3mm thick?
It’s not FR-4, it’s aluminum PCB where insulator is not FR-4 but some thermally conductive dielectric. Total thickness is 3.2mm. Aluminum itself should be around 3mm thick.