New problem for me!

So I just got my PEI sheet put on and now I cant get the bed leveling print off. What’s the best way to get prints off of the bed without scratching the sheet and just curious what are people using to clean the sheet with? Thanks in advance!

The sheet is pretty robust, but you don’t want to cut into it. Be careful if you are using the clam knife, as it is sharp. Parts are most easily removed if they haven’t been left on the printer too long (e.g. a day). If it is really on there, heating up to 90C for ABS will help you get the part off. HIPS is generally very easy to remove. What material?



I list some strategies on removing parts at the bottom of my PEI recap post on the SeeMeCNC forum. Sharp tools will scratch and even chip PEI and once you have an imperfection in the surface, it will telegraph to your parts. The tried and true way to remove a stubborn part is to chill it. That might be a little difficult to do with the integrated heated bed/PEI surface on the TAZ, so try the tip I posted and use electronics freeze spray. You really just need a quick spritz on an edge to get it to start to lift. It will then pop off cleanly.


I am actually using HIPS. Can get most of it off but there are portions that just don’t want to come off. Having a hard time getting the bed to level as well now. For some reason one corner just doesn’t seem to want to go down at all.

While I can’t offer much advice on getting a stuck on print off, I can help future prints from sticking so hard. Raise your Z axis so that the filament barely touches rather than getting presses into the bed as is needed with other methods. You should be able to slide a business card or two sheets of paper between the nozzle and PEI with very little force. Basically twice the distance for most other bed materials. Also, skip the skirt, since they are 1 layer thick they are the hardest to remove.

tmorris9, absolutely! As I state in my PEI thread, PEI is not a magic bullet for first layer sticking problems and, in fact, it is maybe even more important that the Z height is set properly.

Use the single layer cylinder STL file at the bottom of the thread in my signature. Print it, say at .20mm layer height. Then remove it and measure it in a few places. If it is not really close to .2 mm - and in fact if it is much less - you probably found your “sticks too well” problem.

Ok cool, thanks. I’ll try that out and see how it works. Still having a hard time getting the bed level now though. One corner just doesn’t seem to want to lower.

Some thoughts…

Keep in mind that 3 points define a plane. As long as your bed surface is flat, only 3 points are needed to “level” it. So I don’t know what it means to say that 1 of 4 points is too high. Have you taken a straightedge and checked the flatness of the bed?

I use a big wedge knife like the kind you see in the kitchen for dicing things, it’s triangle point and wedge shape easily get under a corner of my prints and then the wedge lifts it, although I don’t use that sheet anymore I use painters tape from Walmart as a cheap replaceable surface.

Its more complex than that, as you have a plane that is sitting on a moveable axis and along that axis it may twist. Statically, yes, but as you move the bed along y, the plane will shift and twist along that axis. While x is marginally adjustable (height above bed) this doesn’t help if your Y is twisted. Its a balancing act to adjust for each corner, repeat until you get the best your system can provide.

My case, the aluminum bed the glass is mounted too is about 2mm twisted from opposite corners when measured from x rails. It could be the y rails twisting, it could be the actually aluminum.

Once I realized this, I spend less time worry about perfection and just get it close as after a layer or two, the print has made its own level reference.

Thanks Makuna, that’s a good point. On deltas, that at least is not an issue. There are a host of other mechanical calibration issues but at least the bed itself is “flat”!

I presumed that the surface (glass / PEI) was suspended above the Y carriage mechanism and attached at 3 points. Then, as long as the Y rails are parallel the bed would be flat. It might be tricky getting the rails parallel but there are ways to do that.

Now some of the firmware auto-calibration and auto-compensation work I’ve seen makes sense. I’ve wondered why you would need compensation (other than backlash) in a Cartesian 3D printer. If the motion of the nozzle across the bed is not completely parallel, then compensation could correct for minor variations.