Would mounting the heated bed on nylon standoffs help? If the platform warps because it heats up, can we stop it getting hot? This would obviously cost some Z
The problem isn’t in the Z-direction, but in the X and Y directions. When the bed heats, it expands, and it expands at a different rate than the platform beneath it. Because the bed is fixed to the platform, it warps. What’s needed is a way to mount the bed so that it can expand sideways but stay in the same rotation and be fixed at a particular point (probably the middle).
That makes sense, I guess that would be why I have been having slightly better results after not tightening the mounting bolts quite so tight.
The intention of the standoffs however would be an attempt to stop heat transferring into the platform so as to prevent it warping
The heated bed is going to change size because it has to heat up. If you keep it fixed in place at it’s mounting point it’s going to warp. Increasing the temperature difference in this case will only increase the relative change of size changes with temperature and make the problem worse.
The other thing to keep in mind is that the heating warpage doesn’t stabilize until about 30 minutes later. I can watch the Z change as the heating element comes on and off. The best solution to this problem is to replace the aluminum frame with something that isn’t affected so much by the heating layer. As I have mentioned a bunch of times already I have purchased some glass filled material that is designed to be an insulator to replace my aluminum bed with. I’ve got to get to this project so I can prove my case.
The metal layer of the top plate that is used has got to be shielding the magnets from affecting the sensor to some degree but maybe not enough. Those magnets are most likely warping the measurement of the sensor to some degree. The question remains, how much does it affect it. That’s part of the reason why physically checking the grid points is necessary.
Someone should put some iron particles on top of the plate and see if some of the field of the magnets is making it through the plate.
I must admit that I did not read all 225 posts in this thread to the letter, so forgive me if I reiterate something… I was wondering why the heated bed is screwed in so tight with so many screws. I’d suppose that the mechanical stress and forces that are intrdoduced from the printing process are not really challenging. Has anyone tried to only fix the heated bed with either only the innermost four screws, or only the outermost four screws instead of the full monty? I guess I’ll give this a try in the next days when I’ve a bit of time at my fingers…
I did actually. But the results weren’t perfect. With the innermost screws it tended to “curl down” at the edges in my case, with the outer screws it bulged a bit at the center.
What I’ve been doing now was to place it on the bed with all screws loose. Heat the bed to standard printing temps and then tighten the screws (not hard, just minimal so it won’t move up or down) starting at the center and then work my way from the inside to the outside.
After looking at the bed level I could then tighten those screws a bit more on those locations where the bed was a bit high. (using the bed level visualizer plugin of octoprint).
But right now I’m on another mission to improve this further: Idea for quick switch bed: feedback requested. I’m thinking that if the printbed is on top of a completely flat surface and supported equally at every the curling up/down of the bed would be a lot less and that leaving some play by not tightening all screws might be better.
But those are just my (basic) experiments, your mileage might vary.
1 Like
I can’t stress enough what a piece of glass will do for you here.
My biggest problem now seems to be my heated bed being very cold on the bottom part of it, which is not really something i can do much about unless i change the heated bed to something else.
I am experimenting with glass directly on the bed vs something in between right now.
I had some adhesvie magnet sheet under the glass originally, in an effort to hold it to the heater (which did not work) and it seems like it had the unintended benefit of being able to distribute the heat even more evenly thru the pane of glass (which im using pyrex)
i next have some kapton tape coming and will repeat with that as well.
I have some additional materials in mind for this as well based on some back and forth and some research, and theres a path i havent tried to go on yet, an insulator for under the bed to reflect the heat back up lost thru the bottom.
i could probably just put the magnet sheet back and be done with it, but in for a penny in for a pound.
Or a piece of MDF underneath the bed 
first experiment with the print sheet on the MDF plate: and only using about 6 screws to attach it to the wood. So my suspicion is already partially confirmed that the support from the bottom for the heated plate is quite important
But beware when looking at all these fancy little graphs of bed levels. This one is the exact same one but with different scales 
but overall, just putting it on a wasteboard, tramming my z-axis parallel with the buildplate has reduced the deviation from highest to lowest point from 0.9mm to 0.26mm
And the bed heats up a bit faster too. Now I’ll see how it will hold up over time.
Regarding the influence of the magnets, I think prusa does calibration with an optimized grid, that won’t probe on positions with magnets.
1 Like
My older 3D printer has a glass bed, and I have been cursing a thousand times about the inconsistent bevaviour on first layer adhesion. I run a print, sticks nice and cool. Run it a second, and third time: All is good. Run it a fourth time: And for my life I cannot get the first layer stick. Nothing changed, same calibration, same GCode, same everything - so after x failed attemps you pick the hairspray and it works then… I don’t know. I’m not so much a fan of glass any more. The snapmaker sheet is the exact opposite experience for me - as long as bed calibration is good - I hardly need a second try ever - first layer is a success basically every time.
I’ve also fitted a glass sheet, but with a magnetic sticker and spring steel pei on top of that. Auto levelling still works, and my first layers are amazing now.
I think it’s partly the flatness of the glass, but also the pei. I don’t know what the Snapmaker print sheets are covered with, but I don’t like it. It also appears they have changed, my initial unit was quite black, but a replacement sheet I bought has a definite blue tinge to it. I think the pei is better than both.
Hello.
what do you think about my bed level?
Thanks
Your scale on the chart is not very useful.
You can’t see the lower portion. I would have it set for ±.2 mm.
You want to be able to see all the hills and valleys.
I cant follow you, It shows the differenz around 0 over the bed, or what do you missing in the chart?
It looks like it is truncating the lower values. You can’t even see 1/3 of the chart. It would be more useful to have a flatter chart which is what you would get with a different z scale.
Cant verify at the moment, will have a look at it, but my digitalprobe shows nearly the same values.