I have a warped heat bed and just received a replacement. Unfortunately it is slightly worse than what I have. Bambu indicated it does not have to be returned so I am going to try to tram it without adding aluminum tape but by removing the high spots on the outside edges of the bed - I said this was crazy. The reason for this approach is the center of the bed is fine, it is the outside edges that curl up. Better to remove a portion of 10% of the bed versus adding tape to 80% of the bed.
If the black top is only a thick continuous magnetic sheet, then removing just portions of it probably won’t reduce the field strength enough to prevent the build plate from sticking to the heat bed.
Anyone been to crazy town and tried this already?
I’d fly cut it on a CNC (I have a CNC). Sanding is going to be very “iffy”, not precise enough to be worth taking a shot at it. And depending on how it’s molded, you may not have enough material there to shave it down and not have the sides peel off.
I’d go and find the flattest thing I could, a bit larger than the build plate, that could take a few 100ºF, and I’d put the plate face down on it in an oven at something around the glass reflow (vitrification) temp for the plastic for a few hours. Then I would let it cool overnight with the oven door closed.
So instead of removing the high material, you’d be “relaxing” the low interior material so it ends up at the same height as the higher edges. No material gets removed.
1 Like
Excellent ideas. I’m going to give the latter one a go and see how it turns out.
Be careful with the temp. Magnets can lose their magnetism if they get too hot.
All magnets lose coercivity (magnetism) as they heat up. It comes back when they cool off. They have to reach their “Curie Temp” to permanently lose magnetism. But none of the common magnetic materials have Curie Temps below around 600ºF. So 200ºF or so isn’t going to be enough to do any harm.
But my assumption was that the magnet is removed first, since the OP’s original idea of sanding down the high spots wouldn’t work with the magnet still there.
1 Like
The initial post says “If the black top is only a thick continuous magnetic sheet, then removing just portions of it probably won’t reduce the field strength enough to prevent the build plate from sticking to the heat bed.”
So the intent is to sand down the magnetic sheet. The warped bed is otherwise useless, so there is nothing to lose by trying.
Agreed, the magnetic sheet would have to be removed before trying to machine the surface. I don’t have a CNC so that is not an optrion for me.
For trying the heat approach, my thought was to
- leave the heat bed mag sheet in place
- put my (flat) glass plate (with its own mag sheet) on top of the bed to help “pull” the low spots up
- invert the plate assembly
- heat the assembly in the oven above 100C and see it that helps flaten the bed
If I loose both mag sheets, no big deal. They can be replaced
I’m not expecting this to work without reaching oven temperatures that might also deform the plastic frame. But I can’t use the replacement bed anyway, so nothing lost.
Ah. I missed the comment implying sanding with the mag-sheet or I’d have specifically suggested removing it. It’ll sand a lot easier than the plastic, but you’d be making it weaker around the perimeter where you want it to be strongest.
A glass sheet will be good provided it can take the heat. 100ºC should not do any harm to the magnets. Ferrous magnets aren’t the strongest, but the Curie Temp is something north of 300ºC.
1 Like
In my case, about 10 - 15mm of the outside edges of the heat bed are too high. My initial thought was if I only remove about 0.2 to 0.3mm from the outside edges of the mag sheet there would still be enough mag field left to still be effective. Especially since they were high points and the center would be pulling the sheet down tight against the outside edges.
We’ll see what the oven does. I’m baking at 150C.
Remember, all of this is just a crazy experiment.
My oven experiment did nothing more than distort the plastic frame. On disassembly I see the heat bed assembly is made up of a 2mm magnetic sheet and a 3mm aluminum plate (with laminated heater sheet). The aluminum plate determines the flatness of the bed. Machining the aluminum plate does not appear to be practical. Here are photos of the components.
Ah. I had no idea what the construction of the bed was like.
Does the plate look “flatter” if it’s not dogged down to the plastic frame?
If you gently twist the plate, does it “oil can” (pop like the metal vacuum top on a bottle of ice tea)?
1 Like
The plastic frame is not rigid enough to effect the Al plate shape.
The bow in the plate is not effected by any hand flexing. It returns to its original shape.
Maybe I’ll post the component breakdown as an FYI for others.
If you’re only talking about a small deflection, the plastic frame is probably capable of affecting the plate. Stuff that looks rigid to the human eye is often a lot more flexible on smaller scales than you’d think.
Given the plate is removable, provided there’s a way to fixture it, it could be fly cut. Any automotive performance company that can deck heads and blocks would probably be able to fixture and deck the Al plate.
But of course, that’s a lot more work than just pestering BBL for a replacement…
2 Likes
Thanks for posting detailed photos. My bed appears to be warped and I can see it visually. I would be interested to see what the machine thinks of bed flatness. Do you know how warped your plate was?
Hm… milling the alu plate could be an solution… i have an bent bed too but actual i cant try it because i cant get further with the ticket it seems they not want to send me one… when the plate ist destroyed after, the whole printer is because there are no spare parts…
My total variation was 0.4mm on the X axis. The lowest spot was in the middle of the plate.
The variation in the center of the plate can be handled by the ABL firmware. This will give you a smooth first layer in the ABL test area but does not make it flat. The latter can cause issues with the bottom part wobbling when placed on a flat surface or tight tolerance parts won’t fit together. And the fit is affected by where the part is printed on the plate.
For my 0.4mm variation in the X axis much of it was outside the ABL test area so there was no adequate compensation. The ABL does not test the left and right 35mm of the print plate on the Xaxis. This limited my max print width in the Xaxis to 200mm.
See the following for more details
https://www.reddit.com/r/BambuLab/comments/13c4yji/suggest_expanding_the_bed_leveling_test_area_in/
Very interesting seeing the disassembly of the heatbed, I am looking to modify mine to attempt improving the flatness, but I am curious how does the black magnetic sheet come off, does it simply just peel up, or is it secured with some sort of adhesive?
I am going to try to see if I can remove the aluminum plate, and sand it down in targeted areas where the plate is high on the machine, and then finish lap it to perfect flatness. If I have success doing this I will try to document my process as I know tons of people would love a permanent fix for the warped bed. I currently have mine shimmed with copper tape and it is better overall, but it isn’t a very smooth transition.
Yes. But the sticker itself gets easily destroyed by that, I had to buy new one.
I doubt you’re going to success with that. The plate is too thin, it warps when heated. It will warp even more if sanded thinner.
Yeah, no, you’re not going to be able to sand down the flex plate to fix this. It’s the plastic “chassis” the plate sits on that needs to be flattened. Sanding the flex plate will only result in a flex plate that cracks or permanently deforms when bent because it’s too thin.
2 Likes
For those claiming the plastic chassis is the problem, it’s not. If you don’t believe that, take apart your bed and it becomes immediately obvious. The plastic chassis is not strong enough to maintain its form when not screwed to the heat plate. It simply is not strong enough to have any meaningful impact of the flatness of the heat plate. The heat plate retains its curve when removed from the plastic chassis and and the steel frame. The amount of force necessary to flex the heat plate is very significant. It hard to get a significant deflection by hand.
The heat plate is stamped out of a larger piece of material and my current theory is poor control in the stamping process is resulting an a “permanent” bend being put in the stamped piece.
1 Like