Page 1 of 1

Question: Directly attach pei to build plate?

Posted: Sat Dec 10, 2016 8:11 am
by innkeeper
I happened to have a chance to play with the new i3 mk2 and the thing that impressed me is that they attached the pei directly to the heater plate/pcb which made the bed heatup times super fast..(like same speed as the hotend) . so that got me to thinking... why cant we do similar on the m2. Has anyone tried this?

some potential issues i can think of:
if the build plate isn't flat
if the build plate warps when heated.
loss of mass in the build area causing unforeseen issues.
possible need to secure the build plate to the spider


Re: Question: Directly attach pei to build plate?

Posted: Sat Dec 10, 2016 3:13 pm
by ednisley
innkeeper wrote:if the build plate isn't flat / warps when heated
Misalignment is the print killer: the need to maintain the Z-axis position closer than ±0.1 mm across the entire expanse of the platform rules out essentially all the cheap, easy, convenient, simple implementations. Conversely, ignoring that tolerance leads to a whole bunch of stuff that produces nothing but headache...

For example, I've been using a hotrod platform that consists of a PCB heater bonded to a borosilicate glass slab. The whole thing mounts to the Y-axis stage through the three alignment screws: no spider, no clamps, just a dead-flat piece of hot glass. This is an early picture, before I finished relocating the platform height switch: ... hermistor/


The guy who built that recently proposed bonding a silicone heater to the M2's original heat spreader (the anodized 1/8 inch = 3 mm aluminum sheet with the Makergear logo), with a slice of PEI bonded to the top. Turns out that his sheet is as flat as he can measure it, while mine has a 0.5 mm trough along the middle; the edges along the +X and -X sides of the sheet have a distinct upward bend. We can't tell how much the sheet will warp when heated, but I'm certain mine won't magically become good enough to be useful.

So, replacing heavy glass with lightweight aluminum will certainly reduce the heating time and, with bit of luck, his platform might actually be flat enough on his M2. However, my version of the same materials definitely won't work on my M2. You can't see that 0.5 mm warp without a known-good straightedge, but that much Z-axis error will destroy any print.

Bonding a PCB heater to a sheet of PEI might work in an M2, but I expect it won't produce stable, accurate results.

Re: Question: Directly attach pei to build plate?

Posted: Fri Dec 23, 2016 6:12 pm
by innkeeper

Though.... still might be workable, at least for me.

I am using a a smoothie board clone with smoothiewear, and I can do mesh compensation on auto leveling/traming so it might still work fine even though its not flat.

in fact, the i3 k2 does the same thing, it does a mesh compensation on the auto levenling / traming.

Perhaps that's the trick letting them do that on such a thin medium, as then it dosn't matter how much the bed warps some (within reason) as long as it is stable during the print.

Re: Question: Directly attach pei to build plate?

Posted: Fri Dec 23, 2016 10:38 pm
by ednisley
innkeeper wrote:it dosn't matter how much the bed warps
That will produce a non-flat bottom surface and, if the platform is tilted, not-quite-perpendicular sides.

I'd say most objects won't show any difference, but anything built to sit flat on a surface will come out wobbly and bonding pieces into a larger gadget won't end well.

Definitely worth trying out, though, just to see how it plays!