Even with an almost perfect calibration, I have a taco heat bed. In addition, I do not understand why the heat bed adjustment threats/ knobs come with applied blue Loctite.
Prints will work, but you get an inconsistency in the first layer on large objects. This leads to bad outer surfaces for the first few layers.
I will need a replacement heat bed.
A variance of 0.6mm is not bad. Try heatsoaking the printer and disable the chamber heater for the first few layers. This thread is having a simmilar issue, althoug not exactly the same: 1 & 2 layer strong problems
IMHO, difference upto 1mm is acceptable for this printer. It will be compensated by firmware when prints.
Please check and tight screws on the rail, on the hotend, on the strain gauge.
And for now maybe something wrong with z-offset when printing from slicer directly. Please try to print the same from printer’s memory.
In general ~0.6mm would be OK, I had problems with >0.9mm only and never get below 0.3 as well.
With the standard print plate (which is one sided) you can try to add tape as described here to flatten it. I did that on one of my older printers and it worked to get the level variation < 0.5.
I used capton tape which is heat resistant!
0.6 variance is a three-layer difference. I don’t understand how you can evaluate this as acceptable. This leads to a very inconsistent first layer. It might even be the reason why you can’t buy a smooth-build plate.
You can clearly see that the compensation algorithm is unable to compensate for this amount of variance; the wall thickness is inconsistent on the first layer.
To replicate print the first layer of any bigger object.
The firmware is “translate” every gcode command (if everything working well), and compensate, more or less, even for a big objects.
But they have a bug if you print from slicer directly. It works OK only if you send to printer and than print from his display.
But on your pictures above i see results most probably due to loosen screw(s) somewhere.
Made this big movable plate stable for a different temeratures - is a challenge. It cannot be 100% straight. In this case, i think, Bambu still havent big printers.
At least you were able to print something… after manually leveling the bed, adding kapton tape, and still warped more than 0.35mm (varies greatly each cal), Creality Hyper PETG will not stick without manual Z offset adjustment. Even then, layer quality is very poor and mesh-leveling is only about half as effective as it should be. Results are different if printed from slicer, or printed from printer. Reverting to old firmware and CrealityPrint5 - no go. Adding M28 to end of start gcode - no go. Frequent firmware crashes / bugs. Retraction issues, CFS/EXT issues… I’m sending it back - not going to wait months for vaporware support on a product not ready for market. Maybe I just got a “bad” one but this thing is useless.
Every PETG needs z-offset adjustment. Creality still not have it in the filament profile - its bad.
I’ve do it manually every PETG print.
But you’re right - product isnt ready for the mass market. See my post with issues.
This is the same 1st layer gcode, printed directly from slicer, and than from printers memory :
This is my today’s test of the new epoxy resin plate.
After printer’s power cycle (but normally i’m not switching it off) it starts first print OK from the slicer directly.
First layer very dependable of the ER plate temperature, better 40 C for Hyper PLA and 60 C for a cheap domestic PETG.
And anyway necessary to add z-offset every print, even for Hyper PLA.
0.050 for Hyper PLA and 0.090 for a PETG and 1s layer is OK
I did everything
-only printing from the printers memory
-leveling with a heat soaked bed. even used the level screws for further optimisation
-checked all screws according Troubleshooting for Layer Compression | Creality Wiki
-did manual z offset
conclusion:
the possible compensation seems to be dependent on the the complexity of the deformation shape
Sure. Not every plate shape can be compensated. IMHO, it depends of smoothness of bending. Sharp bends is a problem.
This is PETG? Did you test the flow rate?
Try to print a cube 10-20mm 100% infill (this is the way to check).
BTW, it can be a fake Creality filament. But almost any PETG possible to print, just slow and with a very different parameters. And better to give the full info about material and temperatures. I cannot wath the tube right now, sorry (Russia).
I reached out to the support for a new plate. I would even pay for one, but local dealers and the EU store do not have any in stock.
Yes, it is PET-G and it is genuine. I played the z-offset and flow rate to the extreme. Besides a new heat plate, the only solution so far is to heavily increase the probing resolution points, which come with an even longer calibration, which is not a long term solution.
By the way, many people are complaining about the slicer, but there is a z-offset configuration
I just received my replacement bed. It’s even worse… The gray pad over the silicon is missing, its bent, and the magnet sheet is not installed. Does Creality any quality checks?! Never had such a frustrating experience in the last few years of 3d printing
My bed was also curved, front to back (taco shape) - I suspected it was warped due to the weight of the CFS that ships inside the printer. However from your experience with a brand new plate, I’m now thinking it may be simply a manufacturing issue. Anyway, I fixed my plate by inserting a shim between the plate and the steel cradle on either side, half way from front to back (mid-way between the bed spring-loaded corner screws). Here are the before and after bed mesh shots:
Bed before adding shims:
Bed after adding shims and levelling with the four corner screws:
So it is possible to correct a curved plate to some degree, but it took me a lot of time and patience to do it.
Nice results. I’m looking for a more “professional” solution.
I’m currently in contact with third-party companies for a thicker and more precise machined bed.
