I have some issues with the print bed of my new SV08. I got it in January 2026 and it looks different from the older ones I saw on printables and forums. See attached picture.
It has factory installed springs at the corners. The openings in the magnetic sheet are also factory installed. The bed ‘floats’ at the corners but not at the center.
The problem is that I cannot lower the corners far enough tightening those corner screws is not possible without breaking something. The center is too low in height. Also the plastic front edge of the plastic printbed case is pushing the printplate up, causing more warping. The plastic edges on left and right are also causing warping when the print plate is not aligned perfectly.
Is there a way to detach the center from the plastic printbed case? I’m hoping to increase the center height and add some more supports under the bed to make the bed more flat.
Why do you need to lower the bed? Back off all 4 corners until the top surface of the magnet sheet is very slightly higher than the plastic edge such that a miss aligned build plate won’t touch. Run quad level, then Z offset.
The springs allow the bed to be mostly isolated from the aluminum frame deep inside the base. It seems Sovol (finally) agrees with me that microscopic twisting occurs in the frame as temperature increases. With the old style bed bolted to the frame twisting of the frame is transferred to the bed. Spring isolation is a good idea. If you crank the screws down until the coils stack together you have turned the spring into a rigid spacer.
I just tried what you suggested. I loosened the screws one full turn and checked that I could push down slightly on the front corner grips and feel the springs move. Then I get a Z difference that is slightly worse; 0.410mm instead of the 0.368 I had before. But like you said, the amount I measured before today varied a lot with the temperature and was often worse. The right-middle part (near 350,175) is usually the deepest part and gives me the most problems with adhesion for the first layer.
The springs are nice but in my case I would like the middle (0,175) to (355,175) to be a bit higher because it seems to me that the bed is bolted in the middle (175,175) to the printbed and not supported enough near the openings near (0,175) and (355,175).
It would make no sense to bolt the middle with springs on the corners.
Go up another turn and run a mesh. Did the “range” get bigger?
Put a strip of packing tape on top of the magnet sheet (under the build plate) and rerun your mesh.
Then add 2 more side by side (don’t overlap the seams) and mesh.
If you are going to run bed temperatures > 60 °C order some Kapton tape to replace the packing tape.
You can also sand down the high spots on the magnet sheet. The tape is reversable while sanding is not.
I tried with an extra turn turn to loosen the corner screws and that made it worse.
I tried adding some thick paper under the print plate, at the right middle and that showed up on the mesh as a flat raised plateau.
Then I started to add alu tape on the spots where Z was the lowest. That helps. I think that with a bit of extra tape and some light sanding of those red parts I can get the height difference below 0.200mm. Will try that over the weekend. And then see how this holds up for higher temperatures.
Once you get it flat you can change it back. If you use [adaptive_mesh] in your start_print macro it will never show you the edges unless you are printing a large print.
So the middle was not actually bolted down to the printbed frame as I thought. You can actually push it down and it will ‘spring’ back up just like at the corners. It just needs a lot more force. I could not see a way to adjust the height in the middle other than by adding more tape there so I continued with that.
I was getting some heightmap measurements that would vary a lot, even if I did not change anything to the plate or bed, for repeated measurements at the same temperature. I saw a comment on youtube from someone claiming that with the new sv08 bed all you had to do was loosen the screws, heatsoak at 100% and adjust until flat enough. That did not help me directly but maybe the heatsoaking stabilized the magnetic sticker and my tape a bit.
After that I changed my bedmesh temperature to 0 to be able to change/measure a bit faster. I thought I was getting closer to 0.2 but the measurements were still varying a lot.
Then I noticed the horizontal_move_z: 5 and changed that to the 2mm I was using before, changed the bedmesh temperature back to 50C and got more reliable measurements. I also made sure to run the front fan of the toolhead at 100% and park the head at Z=200 while I was making changes. The eddy coil is supposed to be more accurate at that height but the Sovol Eddy has no temperature compensation so I tried to improve the accuracy this way and it seemed to work well.
So I got to the point where the Z height difference is now at about 0.2mm and was finally able to print a 350x350 first layer for the first time. Not a great first layer but I think I can probably still improve this.
